A Multi-purpose Realistic Haze Benchmark with Quantifiable Haze Levels and Ground Truth
Extracted source text · arXiv 2206.06427
%% bare_jrnl.tex
%% V1.4b
%% 2015/08/26
%% by Michael Shell
%% see http://www.michaelshell.org/
%% for current contact information.
%%
%% This is a skeleton file demonstrating the use of IEEEtran.cls
%% (requires IEEEtran.cls version 1.8b or later) with an IEEE
%% journal paper.
%%
%% Support sites:
%% http://www.michaelshell.org/tex/ieeetran/
%% http://www.ctan.org/pkg/ieeetran
%% and
%% http://www.ieee.org/
%%*************************************************************************
%% Legal Notice:
%% This code is offered as-is without any warranty either expressed or
%% implied; without even the implied warranty of MERCHANTABILITY or
%% FITNESS FOR A PARTICULAR PURPOSE!
%% User assumes all risk.
%% In no event shall the IEEE or any contributor to this code be liable for
%% any damages or losses, including, but not limited to, incidental,
%% consequential, or any other damages, resulting from the use or misuse
%% of any information contained here.
%%
%% All comments are the opinions of their respective authors and are not
%% necessarily endorsed by the IEEE.
%%
%% This work is distributed under the LaTeX Project Public License (LPPL)
%% ( http://www.latex-project.org/ ) version 1.3, and may be freely used,
%% distributed and modified. A copy of the LPPL, version 1.3, is included
%% in the base LaTeX documentation of all distributions of LaTeX released
%% 2003/12/01 or later.
%% Retain all contribution notices and credits.
%% ** Modified files should be clearly indicated as such, including **
%% ** renaming them and changing author support contact information. **
%%*************************************************************************
% *** Authors should verify (and, if needed, correct) their LaTeX system ***
% *** with the testflow diagnostic prior to trusting their LaTeX platform ***
% *** with production work. The IEEE's font choices and paper sizes can ***
% *** trigger bugs that do not appear when using other class files. *** ***
% The testflow support page is at:
% http://www.michaelshell.org/tex/testflow/
\documentclass[journal]{IEEEtran}
%
% If IEEEtran.cls has not been installed into the LaTeX system files,
% manually specify the path to it like:
% \documentclass[journal]{../sty/IEEEtran}
\usepackage{times}
\usepackage{epsfig}
\usepackage{graphicx}
\usepackage{amsmath}
% *** ALIGNMENT PACKAGES ***
\usepackage{array}
% *** Graphic PACKAGES ***
\usepackage{graphicx}
\usepackage{amssymb}
% *** Multi Row ***
\usepackage{multirow}
\usepackage{adjustbox}
\usepackage{tabularx}
\usepackage{xcolor,colortbl}
\usepackage{booktabs}
\usepackage{xspace}
%------------------------------------------------------------
% *** PDF, URL AND HYPERLINK PACKAGES ***
\usepackage{url}
% *** PSEUDOCODE PACKAGES ***
\usepackage{algorithmic}
\usepackage[ruled, linesnumbered, noend]{algorithm2e}
\usepackage{xcolor}
\usepackage{amssymb}% http://ctan.org/pkg/amssymb
\usepackage{pifont}% http://ctan.org/pkg/pifont
\usepackage{soul}
\usepackage[caption=false]{subfig}
\usepackage{caption}
% \usepackage{subcaption}
\usepackage{enumitem}
\usepackage{footnote}
\usepackage{tablefootnote}
\usepackage{longtable}
\newcommand{\zhenyu}[1]{\textcolor{red}{#1}}
\newcommand{\priya}[1]{\textcolor{blue}{#1}}
\newcommand{\blue}[1]{\textcolor{blue}{#1}}
\newcommand{\red}[1]{\textcolor{red}{#1}}
\newcommand{\cmark}{\ding{51}}%
\newcommand{\xmark}{\ding{55}}%
\makeatletter
\DeclareRobustCommand\onedot{\futurelet\@let@token\@onedot}
\def\@onedot{\ifx\@let@token.\else.\null\fi\xspace}
\def\eg{\emph{e.g}\onedot} \def\Eg{\emph{E.g}\onedot}
\def\ie{\emph{i.e}\onedot} \def\Ie{\emph{I.e}\onedot}
\def\cf{\emph{c.f}\onedot} \def\Cf{\emph{C.f}\onedot}
\def\etc{\emph{etc}\onedot} \def\vs{\emph{vs}\onedot}
\def\wrt{w.r.t\onedot} \def\dof{d.o.f\onedot}
\def\etal{\emph{et al}\onedot}
\let\emptyset\varnothing
\DeclareMathOperator*{\argmin}{argmin}
\DeclareMathOperator*{\argmax}{argmax}
\definecolor{Gray}{gray}{0.9}
\definecolor{GrayLine}{gray}{0.7}
\makeatother
\usepackage[noadjust]{cite}
\usepackage{tcolorbox,varwidth}
\newtcbox{\tightredbox}{ colframe=red,colback=white,arc=0.5pt,boxsep=0pt,left=0pt,right=0pt,top=0pt,bottom=0pt,boxrule=0.5pt,varwidth upper}
\newtcbox{\tightbluebox}{ colframe=blue,colback=white,arc=0.5pt,boxsep=0pt,left=0pt,right=0pt,top=0pt,bottom=0pt,boxrule=0.5pt,varwidth upper}
\usepackage[linewidth=1pt]{mdframed}
\mdfsetup{skipabove=0pt,skipbelow=1pt}
\mdfdefinestyle{RedBox}{%
linecolor=red,
innerlinewidth=3pt,
%roundcorner=20pt,
innertopmargin=0pt,
innerbottommargin=0pt,
innerrightmargin=0pt,
innerleftmargin=0pt,
leftmargin = 0pt,
rightmargin = 0pt,
outermargin = 0pt,
%backgroundcolor=gray!50!white}
}
\mdfdefinestyle{BlueBox}{%
linecolor=blue,
innerlinewidth=3pt,
%roundcorner=20pt,
innertopmargin=0pt,
innerbottommargin=0pt,
innerrightmargin=0pt,
innerleftmargin=0pt,
leftmargin = 0pt,
rightmargin = 0pt,
outermargin = 0pt,
%backgroundcolor=gray!50!white}
}
% Some very useful LaTeX packages include:
% (uncomment the ones you want to load)
% *** MISC UTILITY PACKAGES ***
%
%\usepackage{ifpdf}
% Heiko Oberdiek's ifpdf.sty is very useful if you need conditional
% compilation based on whether the output is pdf or dvi.
% usage:
% \ifpdf
% % pdf code
% \else
% % dvi code
% \fi
% The latest version of ifpdf.sty can be obtained from:
% http://www.ctan.org/pkg/ifpdf
% Also, note that IEEEtran.cls V1.7 and later provides a builtin
% \ifCLASSINFOpdf conditional that works the same way.
% When switching from latex to pdflatex and vice-versa, the compiler may
% have to be run twice to clear warning/error messages.
% *** CITATION PACKAGES ***
%
%\usepackage{cite}
% cite.sty was written by Donald Arseneau
% V1.6 and later of IEEEtran pre-defines the format of the cite.sty package
% \cite{} output to follow that of the IEEE. Loading the cite package will
% result in citation numbers being automatically sorted and properly
% "compressed/ranged". e.g., [1], [9], [2], [7], [5], [6] without using
% cite.sty will become [1], [2], [5]--[7], [9] using cite.sty. cite.sty's
% \cite will automatically add leading space, if needed. Use cite.sty's
% noadjust option (cite.sty V3.8 and later) if you want to turn this off
% such as if a citation ever needs to be enclosed in parenthesis.
% cite.sty is already installed on most LaTeX systems. Be sure and use
% version 5.0 (2009-03-20) and later if using hyperref.sty.
% The latest version can be obtained at:
% http://www.ctan.org/pkg/cite
% The documentation is contained in the cite.sty file itself.
% *** GRAPHICS RELATED PACKAGES ***
%
\ifCLASSINFOpdf
% \usepackage[pdftex]{graphicx}
% declare the path(s) where your graphic files are
% \graphicspath{{../pdf/}{../jpeg/}}
% and their extensions so you won't have to specify these with
% every instance of \includegraphics
% \DeclareGraphicsExtensions{.pdf,.jpeg,.png}
\else
% or other class option (dvipsone, dvipdf, if not using dvips). graphicx
% will default to the driver specified in the system graphics.cfg if no
% driver is specified.
% \usepackage[dvips]{graphicx}
% declare the path(s) where your graphic files are
% \graphicspath{{../eps/}}
% and their extensions so you won't have to specify these with
% every instance of \includegraphics
% \DeclareGraphicsExtensions{.eps}
\fi
% graphicx was written by David Carlisle and Sebastian Rahtz. It is
% required if you want graphics, photos, etc. graphicx.sty is already
% installed on most LaTeX systems. The latest version and documentation
% can be obtained at:
% http://www.ctan.org/pkg/graphicx
% Another good source of documentation is "Using Imported Graphics in
% LaTeX2e" by Keith Reckdahl which can be found at:
% http://www.ctan.org/pkg/epslatex
%
% latex, and pdflatex in dvi mode, support graphics in encapsulated
% postscript (.eps) format. pdflatex in pdf mode supports graphics
% in .pdf, .jpeg, .png and .mps (metapost) formats. Users should ensure
% that all non-photo figures use a vector format (.eps, .pdf, .mps) and
% not a bitmapped formats (.jpeg, .png). The IEEE frowns on bitmapped formats
% which can result in "jaggedy"/blurry rendering of lines and letters as
% well as large increases in file sizes.
%
% You can find documentation about the pdfTeX application at:
% http://www.tug.org/applications/pdftex
% *** MATH PACKAGES ***
%
%\usepackage{amsmath}
% A popular package from the American Mathematical Society that provides
% many useful and powerful commands for dealing with mathematics.
%
% Note that the amsmath package sets \interdisplaylinepenalty to 10000
% thus preventing page breaks from occurring within multiline equations. Use:
%\interdisplaylinepenalty=2500
% after loading amsmath to restore such page breaks as IEEEtran.cls normally
% does. amsmath.sty is already installed on most LaTeX systems. The latest
% version and documentation can be obtained at:
% http://www.ctan.org/pkg/amsmath
% *** SPECIALIZED LIST PACKAGES ***
%
%\usepackage{algorithmic}
% algorithmic.sty was written by Peter Williams and Rogerio Brito.
% This package provides an algorithmic environment fo describing algorithms.
% You can use the algorithmic environment in-text or within a figure
% environment to provide for a floating algorithm. Do NOT use the algorithm
% floating environment provided by algorithm.sty (by the same authors) or
% algorithm2e.sty (by Christophe Fiorio) as the IEEE does not use dedicated
% algorithm float types and packages that provide these will not provide
% correct IEEE style captions. The latest version and documentation of
% algorithmic.sty can be obtained at:
% http://www.ctan.org/pkg/algorithms
% Also of interest may be the (relatively newer and more customizable)
% algorithmicx.sty package by Szasz Janos:
% http://www.ctan.org/pkg/algorithmicx
% *** ALIGNMENT PACKAGES ***
%
%\usepackage{array}
% Frank Mittelbach's and David Carlisle's array.sty patches and improves
% the standard LaTeX2e array and tabular environments to provide better
% appearance and additional user controls. As the default LaTeX2e table
% generation code is lacking to the point of almost being broken with
% respect to the quality of the end results, all users are strongly
% advised to use an enhanced (at the very least that provided by array.sty)
% set of table tools. array.sty is already installed on most systems. The
% latest version and documentation can be obtained at:
% http://www.ctan.org/pkg/array
% IEEEtran contains the IEEEeqnarray family of commands that can be used to
% generate multiline equations as well as matrices, tables, etc., of high
% quality.
% *** SUBFIGURE PACKAGES ***
%\ifCLASSOPTIONcompsoc
% \usepackage[caption=false,font=normalsize,labelfont=sf,textfont=sf]{subfig}
%\else
% \usepackage[caption=false,font=footnotesize]{subfig}
%\fi
% subfig.sty, written by Steven Douglas Cochran, is the modern replacement
% for subfigure.sty, the latter of which is no longer maintained and is
% incompatible with some LaTeX packages including fixltx2e. However,
% subfig.sty requires and automatically loads Axel Sommerfeldt's caption.sty
% which will override IEEEtran.cls' handling of captions and this will result
% in non-IEEE style figure/table captions. To prevent this problem, be sure
% and invoke subfig.sty's "caption=false" package option (available since
% subfig.sty version 1.3, 2005/06/28) as this is will preserve IEEEtran.cls
% handling of captions.
% Note that the Computer Society format requires a larger sans serif font
% than the serif footnote size font used in traditional IEEE formatting
% and thus the need to invoke different subfig.sty package options depending
% on whether compsoc mode has been enabled.
%
% The latest version and documentation of subfig.sty can be obtained at:
% http://www.ctan.org/pkg/subfig
% *** FLOAT PACKAGES ***
%
%\usepackage{fixltx2e}
% fixltx2e, the successor to the earlier fix2col.sty, was written by
% Frank Mittelbach and David Carlisle. This package corrects a few problems
% in the LaTeX2e kernel, the most notable of which is that in current
% LaTeX2e releases, the ordering of single and double column floats is not
% guaranteed to be preserved. Thus, an unpatched LaTeX2e can allow a
% single column figure to be placed prior to an earlier double column
% figure.
% Be aware that LaTeX2e kernels dated 2015 and later have fixltx2e.sty's
% corrections already built into the system in which case a warning will
% be issued if an attempt is made to load fixltx2e.sty as it is no longer
% needed.
% The latest version and documentation can be found at:
% http://www.ctan.org/pkg/fixltx2e
%\usepackage{stfloats}
% stfloats.sty was written by Sigitas Tolusis. This package gives LaTeX2e
% the ability to do double column floats at the bottom of the page as well
% as the top. (e.g., "\begin{figure*}[!b]" is not normally possible in
% LaTeX2e). It also provides a command:
%\fnbelowfloat
% to enable the placement of footnotes below bottom floats (the standard
% LaTeX2e kernel puts them above bottom floats). This is an invasive package
% which rewrites many portions of the LaTeX2e float routines. It may not work
% with other packages that modify the LaTeX2e float routines. The latest
% version and documentation can be obtained at:
% http://www.ctan.org/pkg/stfloats
% Do not use the stfloats baselinefloat ability as the IEEE does not allow
% \baselineskip to stretch. Authors submitting work to the IEEE should note
% that the IEEE rarely uses double column equations and that authors should try
% to avoid such use. Do not be tempted to use the cuted.sty or midfloat.sty
% packages (also by Sigitas Tolusis) as the IEEE does not format its papers in
% such ways.
% Do not attempt to use stfloats with fixltx2e as they are incompatible.
% Instead, use Morten Hogholm'a dblfloatfix which combines the features
% of both fixltx2e and stfloats:
%
% \usepackage{dblfloatfix}
% The latest version can be found at:
% http://www.ctan.org/pkg/dblfloatfix
%\ifCLASSOPTIONcaptionsoff
% \usepackage[nomarkers]{endfloat}
% \let\MYoriglatexcaption\caption
% \renewcommand{\caption}[2][\relax]{\MYoriglatexcaption[#2]{#2}}
%\fi
% endfloat.sty was written by James Darrell McCauley, Jeff Goldberg and
% Axel Sommerfeldt. This package may be useful when used in conjunction with
% IEEEtran.cls' captionsoff option. Some IEEE journals/societies require that
% submissions have lists of figures/tables at the end of the paper and that
% figures/tables without any captions are placed on a page by themselves at
% the end of the document. If needed, the draftcls IEEEtran class option or
% \CLASSINPUTbaselinestretch interface can be used to increase the line
% spacing as well. Be sure and use the nomarkers option of endfloat to
% prevent endfloat from "marking" where the figures would have been placed
% in the text. The two hack lines of code above are a slight modification of
% that suggested by in the endfloat docs (section 8.4.1) to ensure that
% the full captions always appear in the list of figures/tables - even if
% the user used the short optional argument of \caption[]{}.
% IEEE papers do not typically make use of \caption[]'s optional argument,
% so this should not be an issue. A similar trick can be used to disable
% captions of packages such as subfig.sty that lack options to turn off
% the subcaptions:
% For subfig.sty:
% \let\MYorigsubfloat\subfloat
% \renewcommand{\subfloat}[2][\relax]{\MYorigsubfloat[]{#2}}
% However, the above trick will not work if both optional arguments of
% the \subfloat command are used. Furthermore, there needs to be a
% description of each subfigure *somewhere* and endfloat does not add
% subfigure captions to its list of figures. Thus, the best approach is to
% avoid the use of subfigure captions (many IEEE journals avoid them anyway)
% and instead reference/explain all the subfigures within the main caption.
% The latest version of endfloat.sty and its documentation can obtained at:
% http://www.ctan.org/pkg/endfloat
%
% The IEEEtran \ifCLASSOPTIONcaptionsoff conditional can also be used
% later in the document, say, to conditionally put the References on a
% page by themselves.
% *** PDF, URL AND HYPERLINK PACKAGES ***
%
%\usepackage{url}
% url.sty was written by Donald Arseneau. It provides better support for
% handling and breaking URLs. url.sty is already installed on most LaTeX
% systems. The latest version and documentation can be obtained at:
% http://www.ctan.org/pkg/url
% Basically, \url{my_url_here}.
% *** Do not adjust lengths that control margins, column widths, etc. ***
% *** Do not use packages that alter fonts (such as pslatex). ***
% There should be no need to do such things with IEEEtran.cls V1.6 and later.
% (Unless specifically asked to do so by the journal or conference you plan
% to submit to, of course. )
% correct bad hyphenation here
\hyphenation{op-tical net-works semi-conduc-tor}
\begin{document}
%
% paper title
% Titles are generally capitalized except for words such as a, an, and, as,
% at, but, by, for, in, nor, of, on, or, the, to and up, which are usually
% not capitalized unless they are the first or last word of the title.
% Linebreaks \\ can be used within to get better formatting as desired.
% Do not put math or special symbols in the title.
\title{A Multi-purpose Realistic Haze Benchmark with Quantifiable Haze Levels and Ground Truth}
%
%
% author names and IEEE memberships
% note positions of commas and nonbreaking spaces ( ~ ) LaTeX will not break
% a structure at a ~ so this keeps an author's name from being broken across
% two lines.
% use \thanks{} to gain access to the first footnote area
% a separate \thanks must be used for each paragraph as LaTeX2e's \thanks
% was not built to handle multiple paragraphs
%
\author{Priya~Narayanan, %~\IEEEmembership{Member,~IEEE,}
Xin~Hu*, %~\IEEEmembership{Member,~IEEE,}
Zhenyu~Wu*, %~\IEEEmembership{Member,~IEEE,}
Matthew~D~Thielke, %~\IEEEmembership{Member,~IEEE,}
John~G~Rogers, %~\IEEEmembership{Senior Member,~IEEE,}
Andre~V~Harrison, %~\IEEEmembership{Member,~IEEE,}
John A D’Agostino,
James D Brown,
Long~P~Quang, %~\IEEEmembership{Member,~IEEE,}
James R Uplinger, %~\IEEEmembership{Member,~IEEE,}
Heesung~Kwon, %~\IEEEmembership{Member,~IEEE,}
and~Zhangyang~Wang%~\IEEEmembership{Member,~IEEE,}% <-this % stops a space
%\thanks{M. Shell was with the Department
%of Electrical and Computer Engineering, Georgia %Institute of Technology, Atlanta,
%GA, 30332 USA e-mail: (see %http://www.michaelshell.org/contact.html).}% <-this % %stops a space
%\thanks{J. Doe and J. Doe are with Anonymous %University.}% <-this % stops a space
%\thanks{Manuscript received April 19, 2005; revised %August 26, 2015.}}
\thanks{P. Narayanan, M. D. Thielke, J. G. Rogers, A. V. Harrison, L. D. Quang, J. R. Uplinger and H. Kwon are with DEVCOM Army Research Laboratory, Adelphi,
MD, 20783 USA e-mail: priya.narayanan.civ@mail.mil}% <-this % stops a space
\thanks{X. Hu is with Tulane University, New Orleans, LA 70118}% <-this % stops a space
\thanks{Z. Wu is with Texas A \& M University, College Station, TX 77843 }% <-this % stops a space
\thanks{X. Hu and Z. Wu have contributed equally to this paper}% <-this % stops a space
\thanks{Z. Wang is with The University of Texas at Austin, Austin, TX 78712 }% <-this % stops a space
\thanks{J. A. D'Agostino and J.D Brown are with DEVCOM Chemical and Biological Center, Aberdeen Proving Ground, MD 21010 }% <-this % stops a space
\thanks{Manuscript received Sept XX, 2021; revised XX XX, XXX.}}
% note the % following the last \IEEEmembership and also \thanks -
% these prevent an unwanted space from occurring between the last author name
% and the end of the author line. i.e., if you had this:
%
% \author{....lastname \thanks{...} \thanks{...} }
% ^------------^------------^----Do not want these spaces!
%
% a space would be appended to the last name and could cause every name on that
% line to be shifted left slightly. This is one of those "LaTeX things". For
% instance, "\textbf{A} \textbf{B}" will typeset as "A B" not "AB". To get
% "AB" then you have to do: "\textbf{A}\textbf{B}"
% \thanks is no different in this regard, so shield the last } of each \thanks
% that ends a line with a % and do not let a space in before the next \thanks.
% Spaces after \IEEEmembership other than the last one are OK (and needed) as
% you are supposed to have spaces between the names. For what it is worth,
% this is a minor point as most people would not even notice if the said evil
% space somehow managed to creep in.
% The paper headers
\markboth{Journal of \LaTeX\ Class Files,~Vol.~14, No.~8, August~2015}%
{Shell \MakeLowercase{\textit{et al.}}: Bare Demo of IEEEtran.cls for IEEE Journals}
% The only time the second header will appear is for the odd numbered pages
% after the title page when using the twoside option.
%
% *** Note that you probably will NOT want to include the author's ***
% *** name in the headers of peer review papers. ***
% You can use \ifCLASSOPTIONpeerreview for conditional compilation here if
% you desire.
% If you want to put a publisher's ID mark on the page you can do it like
% this:
%\IEEEpubid{0000--0000/00\$00.00~\copyright~2015 IEEE}
% Remember, if you use this you must call \IEEEpubidadjcol in the second
% column for its text to clear the IEEEpubid mark.
% use for special paper notices
%\IEEEspecialpapernotice{(Invited Paper)}
% make the title area
\maketitle
\input{chapters/0_Abstract}
% Note that keywords are not normally used for peerreview papers.
\begin{IEEEkeywords}
Degraded Visual Environment, Dehazing, UAV, Object Detection, visual artefacts, benchmarking
\end{IEEEkeywords}
% For peer review papers, you can put extra information on the cover
% page as needed:
% \ifCLASSOPTIONpeerreview
% \begin{center} \bfseries EDICS Category: 3-BBND \end{center}
% \fi
%
% For peerreview papers, this IEEEtran command inserts a page break and
% creates the second title. It will be ignored for other modes.
\IEEEpeerreviewmaketitle
\input{chapters/1_Introduction}
\input{chapters/2_PrevisouWork}
\input{chapters/3_A2I2-Haze}
\input{chapters/4_QuantitativeEvaluation}
\input{chapters/5_Discussion}
\input{chapters/6_Conclusion}
% if have a single appendix:
%\appendix[Proof of the Zonklar Equations]
% or
%\appendix % for no appendix heading
% do not use \section anymore after \appendix, only \section*
% is possibly needed
% use appendices with more than one appendix
% then use \section to start each appendix
% you must declare a \section before using any
% \subsection or using \label (\appendices by itself
% starts a section numbered zero.)
%
%\appendices
%\section{}
%Appendix one text goes here.
% you can choose not to have a title for an appendix
% if you want by leaving the argument blank
%\section{}
%Appendix two text goes here.
% use section* for acknowledgment
\section*{Acknowledgment}
The authors would like to thank US Army Artificial Innovation Institute (A2I2) for funding the data collection and annotation, and financially supporting graduate students at Texas A \& M University for this project. The authors also acknowledges experimentation support from DEVCOM Chemical \& Biological Center (CBC) during the Data collection process.
% Can use something like this to put references on a page
% by themselves when using endfloat and the captionsoff option.
\ifCLASSOPTIONcaptionsoff
\newpage
\fi
% trigger a \newpage just before the given reference
% number - used to balance the columns on the last page
% adjust value as needed - may need to be readjusted if
% the document is modified later
%\IEEEtriggeratref{8}
% The "triggered" command can be changed if desired:
%\IEEEtriggercmd{\enlargethispage{-5in}}
% references section
% can use a bibliography generated by BibTeX as a .bbl file
% BibTeX documentation can be easily obtained at:
% http://mirror.ctan.org/biblio/bibtex/contrib/doc/
% The IEEEtran BibTeX style support page is at:
% http://www.michaelshell.org/tex/ieeetran/bibtex/
%\bibliographystyle{IEEEtran}
% argument is your BibTeX string definitions and bibliography database(s)
%\bibliography{IEEEabrv,../bib/paper}
%
% <OR> manually copy in the resultant .bbl file
% set second argument of \begin to the number of references
% (used to reserve space for the reference number labels box)
\bibliographystyle{IEEEtran}
\bibliography{IEEEabrv}
% {\small
% \input{bare_jrnl.bbl}
% }
\input{chapters/7_Biography}
% that's all folks
\end{document}
\begin{abstract}
Imagery collected from outdoor visual environments is often degraded due to the presence of dense smoke or haze. A key challenge for research in scene understanding in these degraded visual environments (DVE) is the lack of representative benchmark datasets. These datasets are required to evaluate state-of-the-art vision algorithms (e.g., detection and tracking) in degraded settings.
In this paper, we address some of these limitations by introducing the first realistic hazy image benchmark, from both aerial and ground view, with paired haze-free images, and in-situ haze density measurements. This dataset was produced in a controlled environment with professional smoke generating machines that covered the entire scene, and consists of images captured from the perspective of both an unmanned aerial vehicle (UAV) and an unmanned ground vehicle (UGV). We also evaluate a set of representative state-of-the-art dehazing approaches as well as object detectors on the dataset. The full dataset presented in this paper, including the ground truth object classification bounding boxes and haze density measurements, is provided for the community to evaluate their algorithms at: \textit{\url{https://a2i2-archangel.vision}}. A subset of this dataset has been used for the ``Object Detection in Haze'' Track of CVPR UG2 2022 challenge at \textit{\url{http://cvpr2022.ug2challenge.org/track1.html}}.
\end{abstract}
\begin{figure*}[!htb]
\captionsetup[subfigure]{labelformat=empty,justification=centering,farskip=1pt,captionskip=1pt}
\centering
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/1.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/11.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/3.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/4.jpg}
}
\\
\vspace{+0.5em}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/5.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/10.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/7.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/8.jpg}
}
\vspace{+0.5em}
\\
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/13.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/14.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/15.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/16.jpg}
}
\\
\vspace{+0.5em}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/image00002.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/image00025.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/image00131.jpg}
}
\subfloat
{
\includegraphics[width=0.235\linewidth]{figures/paired_images/image00223.jpg}
}
\caption{Examples of paired hazy and haze-free images in A2I2-UAV (top three rows) and A2I2-UGV (bottom one row) from A2I2-Haze.
}
\label{fig:a2i2_teaser}
\end{figure*}
\section{Introduction }
\IEEEPARstart{S}{cene} understanding in outdoor environments for applications such as intelligence, surveillance, and reconnaissance (ISR), and autonomous vehicle navigation is extremely challenging in the presence of smoke, and other adverse weather conditions due to haze, fog, and mist. These atmospheric phenomena with smoke particles or microscopic water droplets significantly interfere with the operations of onboard vision systems, often resulting in imagery with non-linear noise, blur, reduced contrast levels, and color dimming issues. These visual artifacts, generated from uncontrolled and potentially dynamic outdoor environments or other DVE effects, pose major challenges in many components of semantic scene understanding, including image enhancement, image restoration, object localization, and object classification.
To address these challenges, a key requirement is benchmarks to accurately evaluate the performance of these algorithms relative to different quantifiable haze levels. This is beyond the reach and scope of most of the existing curated haze datasets such as RESIDE~\cite{li2019benchmarking}, NH-Haze~\cite{NH-Haze_2020}, and REVIDE~\cite{REVIDE}. Furthermore, these datasets are inadequate to quantitatively and fairly compare computer vision algorithms on hazy vs. haze-free imagery on a specific scene. They cannot isolate and measure the effect of haze and have a shortfall in scene object diversity (Discussed in Section II).
In this study, we leverage the US Army’s unique capability to produce and measure smoke/obscurant to generate haze in a controlled fashion. We collect imagery of target objects such as civilian vehicles, mannequins, and man-made obstacles from UAVs and UGVs. We also collect metadata such as altitude and local haze density from moving and stationary sensors. We then develop an image dataset with metadata for realistic, accurate, and fine-grained algorithm evaluation in hazy DVEs. In summary, the contributions of this paper are as follows:
a) We present A2I2-Haze, the first realistic haze dataset with in-situ smoke measurement aligned to aerial and ground imagery. This multi-purpose dataset has paired hazy and haze-free imagery to allow fine-grained evaluation of low-level vision (dehazing) and high-level vision (detection) tasks. Exemplar images are shown in Fig.~\ref{fig:a2i2_teaser}.
b) We conduct a comprehensive study and evaluation on state-of-the-art single image dehazing and object detection algorithms using this non-synthetic benchmark dataset.
\section{Previous Work }
This section summarizes some of the key findings from previous DVE studies. In subsection A, we provide an overview of the hazy DVE datasets that are publicly available for benchmarking the performance of computer vision algorithms. We then survey current state-of-the-art single-image dehazing methods and object detection techniques used for evaluating DVE datasets in subsections B and C, respectively.
%
\subsection{Haze Datasets}
Several of the currently available hazy DVE datasets (Table 1), such as 3R~\cite{zhang2020nighttime}, HazeRD~\cite{Zhang:HazeRD:ICIP17b}, 4K~\cite{4K_Xiao_2022} and RESIDE~\cite{li2019benchmarking} were synthetically generated using scattering models to simulate hazy conditions. However, while these simulation techniques are useful in generating large-scale training datasets comprising both hazy and reference haze-free images, they often use unrealistic parameters and assumptions such as homogeneity. Further, while synthetically simulating haze ensures all other conditions in the scene are preserved, the technique still poses challenges such as transferring the knowledge to real-world target domains.
More recently, a few real-world hazy datasets such as MRFID~\cite{liu2020end}, BeDDE~\cite{zhao2020dehazing}, Dense-Haze~\cite{ancuti2019dense}, O-Haze~\cite{O-HAZE_2018}, I-Haze~\cite{I-HAZE_2018}, REVIDE~\cite{REVIDE}, and NH-Haze~\cite{NH-Haze_2020} have been published, with images captured in the presence of haze that is generated using professional-grade haze machines. Among the six datasets listed above, NH-Haze is the most representative of a realistic haze scene with non-homogeneous hazy and haze-free pairs. MRFID, BeDDE, Dense-Haze, NH-Haze and O-Haze are datasets with outdoor scenes that make them relevant to ISR and autonomous UGV applications. However, these datasets also pose several limitations. The smoke generators and the sensors for data collection in the MRFID, BeDDE, Dense-Haze, and (NH, O, and I)-Haze datasets are typically mounted in a fixed location and do not capture the spatial variability of the scene. Furthermore, all these datasets are limited to ground view imagery. Moreover, they do not provide a measurement of haze or smoke transmissibility for training and evaluation. The main distinction between our A2I2-Haze dataset from previous efforts is that our dataset provides ground and aerial imagery with haze-free reference images taken from mobile sensors in an outdoor environment. We also provide highly synchronized qualitative and quantitative measurements of smoke transmissibility, using human assessments and in-situ measurements from ground sensors. This will be discussed in detail in subsequent sections.
\begin{table}[ht]
\caption{Properties of A2I2-Haze relative to other public dehazing benchmarks. A/G stands for aerial-/ground-view. In/Out stands for in/outdoor. NH stands for non-homogeneous. HM stands for haze measurement. Details of A2I2-Haze are in the first paragraph of Section IV.}
\centering
\resizebox{\columnwidth}{!}{
\begin{tabular}{c|c|c|c|c|c|c}
\hline
\multirow{2}{*}{Datasets} & \multicolumn{6}{c}{Attributes} \\
\cline{2-7}
& A/G & In/Out & NH & Syn/Real & \#Images & HM \\
\hline
3R~\cite{zhang2020nighttime} & G & O & & S & $2,750$ & \\
HazeRD~\cite{Zhang:HazeRD:ICIP17b} & G & O & & S & $33$ & \\
O-Haze~\cite{O-HAZE_2018} & G & O & & R & $45$ & \\
I-Haze~\cite{I-HAZE_2018} & G & I & & R & $35$ & \\
4K~\cite{4K_Xiao_2022} & G & O & & S & $10,000$ & \\
REVIDE~\cite{REVIDE} & G & I & & R & 1982 & \\
RESIDE~\cite{li2019benchmarking} & G & I+O & & S+R & $100,076$ & \\
BeDDE~\cite{zhao2020dehazing} & G & O & & R & $208$ & \\
MRFID~\cite{liu2020end} & G & O & & R & $800$\tablefootnote{MRFID contains images from 200 clear outdoor scenes. For each of the clear images, there are 4 images of the same scene containing different densities.} & \\
Dense-Haze~\cite{ancuti2019dense} & G & O & & R & $33$ & \\
NH-Haze~\cite{NH-Haze_2020} & G & O & \cmark & R & $55$ & \\
{\bf A2I2-Haze} & {\bf A+G} & {\bf O} & {\bf \cmark} & {\bf R} & {\bf 1,033} & {\bf \cmark} \\
\hline
\end{tabular}
}
\label{table:nonlin}
\end{table}
\subsection{Single Image Dehazing}
Based on \textit{Atmospheric scattering model} (ASM)~\cite{mccartney1976optics,nayar1999vision,narasimhan2003contrast}, a hazy image $I$ can be represented as:
\begin{equation}
I(x)=t(x)J(x)+(1-t(x))A,
\end{equation}
where $J$, $t$, and $A$ denote the latent haze-free image, transmission map, and
global atmospheric light, respectively.
Dehazing using ASM involves estimating the transmission map $t(x)$ and global atmospheric light $A$. Existing approaches for dehazing can be broadly categorized as prior-based methods and learning-based methods.
\subsubsection{Prior-based Methods}
Dehazing methods based on priors~\cite{peng2007image,rao2005algorithms,berman2016non,fattal2008single,fattal2014dehazing,he2010single,tan2008visibility,zhu2015fast} first estimate transmission maps by exploiting the statistical properties of clean images and then obtain dehazed results using the scattering model.
%
Tan~\etal~\cite{tan2008visibility} proposed an adaptive contrast enhancement method for haze removal by maximizing the local contrast of hazy images.
%
He~\etal~\cite{he2010single} put forward an approach with a dark channel prior (DCP) that assumes the existence of at least one channel for every pixel whose value is close to zero.
%
Zhu~\etal~\cite{zhu2015fast} proposed a color attenuation prior for haze removal by estimating the scene depth using a linear model.
%
Fattal~\cite{fattal2014dehazing} proposed a color-line prior for the transmission map estimation by exploiting the regularity in natural images wherein small image patches lay in a one-dimensional distribution in the RGB color space.
%observed that pixels of image patches lay in a one-dimensional distribution in RGB color space and proposed a color-line prior for the transmission map estimation.
%
Berman~\etal~\cite{berman2016non} proposed a non-local prior based on a key observation that pixels in a given cluster are often non-local. Thus, colors of a haze-free image could be approximated by a few hundred distinct colors.
%
Despite some promising results, prior-based approaches have limitations in performance because some of the strong assumptions of these hand-crafted priors often do not hold in a real-world environment.
\subsubsection{Learning-based Methods}
With the availability of large-scale paired data and powerful CNNs, learning-based dehazing methods have become popular in recent years.
%
MSCNN~\cite{ren2016single} estimated transmission map of the hazy images in a coarse-to-fine manner, where the coarse-scale net produced a holistic map based on the whole image and the fine-scale net refined it locally.
%
DehazeNet~\cite{cai2016dehazenet} proposed an end-to-end network that learned and estimated the mapping relations between hazy image patches and their corresponding medium transmissions.
%
DCPDN~\cite{zhang2018densely} embedded the atmospheric scattering model into the network so that it could jointly learn to estimate the transmission map, atmospheric light, and dehazing.
%
Dehaze-cGAN~\cite{li2018single} estimated the clean image based on an encoder-decoder-based conditional generative adversarial network (cGAN). It also introduced the VGG features and $\ell_1$ regularized gradient prior to improving realism.
%
AOD-Net~\cite{li2017aod} recovered hazy images by reformulating the physical scattering model and developing a lightweight CNN to extract clean images. This approach has been extended to an end-to-end video dehazing and detection network~\cite{li2018end}.
%
GFN~\cite{ren2018gated} introduced a multi-scale gated fusion end-to-end encoder-decoder network. This network obtains dehazed images by gating the important features of three inputs derived from the original hazy image by applying White Balance (WB), Contrast Enhancing (CE), and Gamma Correction (GC).
%
EPDN~\cite{qu2019enhanced} formulated the dehazing task as an image-to-image translation problem and proposed an enhanced pix2pix network to solve it. This work also introduced Perceptual Index (PI) as a metric to evaluate the dehazing quality from the perceptual perspective.
%
MSBDN~\cite{dong2020multi} used a boosting strategy and error feedback for progressive restoration.
%
4KDehazing~\cite{zheng2021ultra} proposed an ultra-high-definition image dehazing method via multi-guided bilateral learning. CNNs were used to build an affine bilateral grid that maintained detailed edges and textures.
%
Domain-invariant single image dehazing~\cite{shyam2021towards} recovered images affected by unknown haze distribution using spatially-aware channel attention to ensure feature enhancement and consistency between recovered and neighboring pixels.
%
MSRL-DehazeNet~\cite{yeh2019multi} embedded a multi-scale deep residual learning into an image decompoosition-guided framework, for single image haze removal.
%
DM$^2$F-Net~\cite{deng2019deep} presented a deep multi-model fusion network to attentively integrate multiple models to separate layers
for single-image dehazing.
%
LAP-Net~\cite{li2019lap} learned different levels of haze with different supervisions so that each stage could focus on a region with specific haze level.
%
Despite the improved performance, learning-based methods also have limitations. They require a large amount of data (often paired hazy and haze-free images), and they don't have debugging capabilities on failure cases.
\subsection{Object Detection}
In the deep learning era, object detectors can be grouped into two genres: ``two-stage detectors'' and ``one-stage detectors''.
%
One-stage detectors classify and localize objects in a single-shot for dense prediction. Two-stage detectors have a region proposal module for sparse prediction. Compared to one-stage detectors, two-stage detectors usually achieve better accuracy but lower speed. RCNN series (RCNN~\cite{girshick2014rich}, Fast-RCNN~\cite{girshick2015fast}, Faster-RCNN~\cite{ren2016faster}, R-FCN~\cite{dai2016r}, and Mask-RCNN~\cite{he2017mask}) are the most representative of two-stage detectors. YOLO~\cite{redmon2016you,redmon2017yolo9000,redmon2018yolov3,bochkovskiy2020yolov4,ge2021yolox}, SSD~\cite{liu2016ssd}, RetinaNet~\cite{lin2017focal}, CenterNet~\cite{zhou2019objects}, and FCOS~\cite{tian2019fcos} are some of the state-of-the-art one-stage detectors.
%
\paragraph{Architecture} The anatomy of an object detector includes a backbone, a neck, and a head. The backbone~\cite{he2016deep,howard2017mobilenets} serves as a feature extractor that starts from low-level structures to high-level semantics. The neck, composed of several bottom-up and top-down paths, is adopted to collect feature maps from different stages (scales). The most representative necks include FPN~\cite{lin2017feature}, PAN~\cite{liu2018path}, BiFPN~\cite{tan2020efficientdet}, and NAS-FPN~\cite{ghiasi2019fpn}. The head makes predictions in a multi-scale fashion. Head could either decouples object localization and classification (in two-stage detectors) or simultaneously makes the predictions for localization and classification (in one-stage detectors). In the head, anchors typically give the objects' prior location, shape and size. Recently, anchor-free one-stage detectors such as CenterNet~\cite{zhou2019objects}, FCOS~\cite{tian2019fcos}, and YOLOX~\cite{ge2021yolox} have gained popularity.
\paragraph{Label Assignment} Label assignment defines classification and regression targets for each anchor/grid cell. Traditional assigning strategies utilize local-view information, such as Intersection-over-Union (IoU)~\cite{ren2016faster,redmon2016you} or Centerness~\cite{tian2019fcos}. DeTR~\cite{carion2020end} is the first work that revisits label assignments from a global view, by considering one-to-one assignments using the Hungarian algorithm. To obtain the optimal global assignment under the one-to-many situation, OTA~\cite{ge2021ota} formulates label assignment as an Optimal Transport problem that defines each ground truth (together with the background) as the supplier and each anchor/grid cell as the demander. Thus, the best assignment could be obtained by solving the optimal transport in a Linear Programming form.
\begin{figure}
\centering
\includegraphics[width=1\linewidth]{figures/smoke_generator.png}
\caption{M56E1 smoke-generating system for producing large scale obscuration.}
\label{fig:smoke_generator}
\end{figure}
\begin{figure}[!htb]
\captionsetup[subfigure]{labelformat=empty,justification=centering,farskip=1pt,captionskip=1pt}
\centering
\subfloat{
\includegraphics[width=0.95\linewidth]{figures/data_collection/layout.png}
}
% \subfloat
% {
% \includegraphics[width=0.475\linewidth]{figures/data_collection/data_collection1.JPG}
% }
\\
\subfloat
{
\includegraphics[width=0.475\linewidth]{figures/data_collection/data_collection2.JPG}
}
% \subfloat
% {
% \includegraphics[width=0.475\linewidth]{figures/data_collection/data_collection3.JPG}
% }
\subfloat
{
\includegraphics[width=0.475\linewidth]{figures/data_collection/target_board.png}
}
% \subfloat
% {
% \includegraphics[width=0.475\linewidth]{figures/data_collection4.JPG}
% }
\caption{Example images demonstrating the data collection process. The top figure shows a bird-view layout of the testing site. The yellow circles show target contrast boards, and the red circle shows an infrared and a visible camera. The bottom two figures show target contrast boards (the black and white pylons) used for local haze density measurement.}
\label{fig:data_collection}
\end{figure}
\section{A2I2-Haze Dataset}
The A2I2-Haze Dataset consists of paired haze and haze-free aerial and ground imagery acquired by a small UAS and UGV, respectively. The data set is synchronized with in-situ smoke measurements as well as altitude data acquired from the flight controller. A2I2-Haze were captured in two separate settings – a) grass field as the background, and b) ground with concrete slabs as the background. For the aerial dataset, after each trial, target objects were re-positioned to generate different configuration of objects. The images with good pairing were then selected from the trials. However, the annotated ground dataset is only available from one configuration of target objects at this time. This section provides a detailed description of the data collection procedure and the post-processing pipeline.
\subsection{Data Collection}
Data collection was conducted at the DEVCOM Chemical and Biological Center's (CBC) M-field test range utilizing the world's most comprehensive obscurant generation facility and assessment technologies to measure the smoke concentration accurately. The facility has a platoon of six M56E1 smoke-generating systems (shown in Fig.~\ref{fig:smoke_generator}) that provide large area obscuration by disseminating obscurants either simultaneously or separately while stationary or mobile. The current onboard obscuration capabilities include visual obscuration using fog oil, infrared obscuration using graphite, and radar obscuration using carbon fiber. These obscurants thus can degrade the perception capabilities of threat weapons and Reconnaissance, Intelligence, Surveillance, Targeting, and Acquisition (RISTA) operating in the Visual, Infra-Red (IR), and Millimeter Wave (MMW) portions of the electromagnetic spectrum. Specifically, the obscurant materials are designed to effectively absorb and scatter energy, thus making target acquisition difficult.
As part of the Phase 1 DVE data collection, we obtained a wide range of imagery of targets in the DEVCOM CBC M-Field test area using visual obscurants generated by using fog oil in the M56E1 Smoke-Generating System. The target objects included vehicles, mannequins, and man-made obstacles encountered during UGV maneuvers, such as traffic cones, barriers, and barricades. The ground truth for smoke concentration was measured using laser-based transmissometers. These instruments measured transmittance through the smoke cloud at a $625$nm wavelength and provided a quantitative measure of the clouds' effectiveness in attenuating visible light. The transmissometers' light sources were Z-Laser, S3 series diode modules (Edmund Optics; Barrington, NJ) with output wavelength and power of $625$nm and $5$mW, respectively.
The laser power received at the detector was measured before obscuration, providing a baseline to compare the effect of the fog oil obscurant. During the series of trials, the transmittance was reduced from $100\%$ to below $1\%$, providing a range of obscurant concentrations to correlate with sensor performance. This variation in transmittance (and thus obscurant concentration) is critical as real-world scenarios will have varying haze concentrations. All measurements were time-synchronized with the UAV and UGV sensors to relate the obscuration properties to the sensor's response.
In addition to the transmissometer-based obscuration measurement, we used target boards, as shown in Fig.~\ref{fig:data_collection}, as the second technique for haze measurement. The black and white pylons in the bottom two figures show target contrast boards used for haze density local measurement.
\subsubsection{UAV Dataset}
Aerial video of the DVE releases was captured by Deep Purple 3 (DP3), a custom UAS developed and operated by DEVCOM CBC. DP3 was primarily designed to carry sensor payloads weighing up to 6 lbs using the Array Configured of Remote Networked Sensors (ACoRNS) interface system. In addition to sensor payloads, DP3 can also be equipped with a variety of camera systems. For the A2I2-Haze data collection, a pair of cameras were mounted in the nosecone to capture both longwave Infrared (LWIR) and visible spectrum. DP3 was equipped with a FLIR Boson 640 ($95^\circ$ FOV) for LWIR data capture. Calibrated visible spectrum data was captured using an ELP-USBFHD01M-L21 with a 2.1mm lens, approximately $120^\circ$ field of view, and an OV2710 sensor. Cameras mounted on the nosecone could be rotated to face $0^\circ$, $45^\circ$, $90^\circ$ relative to the body of the UAS, with $90^\circ$ being a nadir pointing the camera, and $0^\circ$ in line with the nose of the UAS. For the A2I2 collection, cameras were clocked at either $90^\circ$ or $45^\circ$. The UAS is flying in a lawn-mowing pattern based on pre-programmed GPS coordinates (aka, waypoints). The flight plan for the collection was defined using a survey grid in Mission Planner to generate a cross-hatch pattern over a predefined area. The area was selected using an arbitrary rectangle within the Mission Planner, with the bounds of the rectangle selected to keep the area of interest within view of the cameras as much as possible over a range of altitudes. The survey grid was cross-hatched with a 10m lane spacing. In order to maintain consistency between altitudes, the survey grid was copied, and the altitude was increased in 5m increments from 15-50m. UAS was commanded to face toward the center of the grid at all points during its flight. The GPS coordinate for the center of the grid was determined by the UAS by manually placing the UAS at the point of interest and using its GPS to determine the center of the grid.
\subsubsection{UGV Dataset}\label{sec:ugv_dataset_generation}
In addition to the aerial vehicle dataset collection described above, a ground vehicle was also used to capture a second viewpoint from inside the haze effect on the surface. The ground vehicle used to collect this dataset is a Clearpath Husky mobile robot which is equipped with a variety of cameras and lidar sensing modalities. The visual dataset for this paper was collected from the color image from the left camera of a Carnegie Robotics MultisenseSL stereo + lidar sensor, which was mounted in a forward-facing position at the front of the robot at the height of 0.6 meters. This sensor package also generated a frame-synchronized second monochrome image from the right camera, a depth image, and a lidar point cloud from the included Hokuyo 2D lidar scanner, which is actuated through rotation about the forward axis to cover a 3D hemispherical volume. A FLIR Boson infrared camera was also used to capture thermal images of the scene. The robot also collected point clouds from an Ouster OS1-64 lidar which is used with a robot mapping engine based on OmniMapper~\cite{trevor2014omnimapper} to localize the UGV in a global frame of reference. The trajectory of the UGV was teleoperated from a remote location via live sensor feedback to collect these datasets. In this paper, only the visual imagery is analyzed; however, the analysis of additional sensor modalities is planned for future work and can be made available to other interested researchers.
\subsection{Data Augmentation and Synchronization }
The raw data collected by sensors were subjected to a series of data augmentation and synchronization procedures in order to enhance their quality. The high mobility of UAV-mounted cameras has brought additional challenges compared to traditional datasets, such as variations in altitude, view angles, and weather conditions. NDFT~\cite{wu2019delving} named these variations as UAV-specific nuisances, which constitute a large number of fine-grained domains. NDFT shows that these nuisances can be used to train a cross-domain object detector that stays robust to many fine-grained domains. As part of A2I2-Haze, we collected these UAV-specific nuisances as metadata that are synchronized with UAV's clock.
\begin{figure}[!htb]
\captionsetup[subfigure]{labelformat=empty,justification=centering,farskip=1pt,captionskip=1pt}
\centering
\subfloat
{
\includegraphics[width=0.45\linewidth]{figures/metadata_sync/uav_baro_transm.png}
}
\subfloat{
\includegraphics[width=0.45\linewidth]{figures/metadata_sync/uav_contrast.png}
}
\caption{UAV visual data synchronized with collected metadata. The left figure shows the visual data synchronized with the barometer and transmissometer metadata. The barometer records the height data of the UAV, and the transmissometer records the laser transmission rate, which measures the haze density. The right figure shows the synchronization of visual data with contrast on three target boards, the second technique that measures the haze density.}
\label{fig:sync_uav}
\end{figure}
\begin{figure}[!htb]
\captionsetup[subfigure]{labelformat=empty,justification=centering,farskip=1pt,captionskip=1pt}
\centering
\includegraphics[width=0.8\linewidth]{figures/metadata_sync/ugv_transm.png}
\caption{UGV visual data synchronized with contrast on target board, which is the second technique that measures the haze density.}
\label{fig:sync_ugv}
\end{figure}
Fig.~\ref{fig:sync_uav} shows the UAV visual data synchronized to the metadata, including the altitude measured by UAV's barometer, and haze density measured by two independent techniques (section III A)- a) using laser transmission rate on a transmissometer and b) using visual contrasts on three target boards. Similar techniques were used for synchronizing the UGV dataset to haze density measurement, as shown in Fig.~\ref{fig:sync_ugv}.
% \subsection{Haze/Haze-Free Pairing for UAV Dataset (\zhenyu{Zhenyu})}
% Subsection text here.
\subsection{Labeling and pairing of UAV Dataset}
Quantitative evaluation of low-level dehazing and high-level detection tasks require paired (hazy and haze-free) images with annotated target objects.
The detection task requires accurate data curation using 2D bounding box annotation tools to define regions of interest. Annotation was performed by external annotators from Engineering and Computer Simulations (ECS) using their labeling platform. Ten object classes were provided as an ontology for labeling. The objects classes were: Sedan, Van, Pickup Truck, Utility Task Vehicle (UTV), Mannequin, Unmanned Ground Vehicle (UGV), Barrel, Jersey Barrier, Aluminum Truss, and Red backpacks. The annotators provided rectangular bounding boxes for each object. They also provided the subjective haze level for each bounding box as light, medium, or heavy.
Dehazing task requires paired hazy and haze-free images. Due to UAV's high flying altitude and rapid movement speed, we sought scene-level pairing rather than pixel-level pairing. We proposed a ``Coarse-to-Fine Matching'' strategy to achieve scene-level pairing. Given two video sequences (hazy and haze-free), we first cut each video into short clips of 2 seconds. We then manually selected hazy and haze-free clips that were similar at the scene level. Lastly, for each selected clip pair ($V_s$ and $V_t$), we used Alg.~\ref{algo:coarse2fine} to find the best pair-able frames with the least amount of relative translation or rotation. Since $V_s$ and $V_t$ were describing the same scene, we hypothesize that the homography matrix (computed from keypoint matching by LoFTR~\cite{sun2021loftr}) is mostly composed of rotational and translational transformation.
The distance of the homography matrix to an identity matrix served as a metric that measures scene-level similarity. We assume that the best pair-able frames have the smallest distance. We manually check the returned paired hazy and haze-free images by the proposed ``Coarse-to-Fine Matching'' strategy. If it fails, we manually find the best pair-able images given the selected hazy and haze-free clips that are similar at the scene level. Such a final human intervention could guarantee the correctness even in the presence of dense haze.
\begin{algorithm}
\caption{Coarse-to-Fine Matching}\label{algo:coarse2fine}
\SetAlgoLined
$N,M \gets \text{Size}(V_s), \text{Size}(V_t)$ \\
$d_{min} \gets \infty$\\
\For {$i \gets 1$ \KwTo $N$} {
\For{$j \gets 1$ \KwTo $M$}{
$I_s, I_t \gets V_s[i], V_t[j]$ \\
\tcp{Keypoints Matching}
$p_s, p_t \gets {\small \text{LoFTR}(I_s, I_t)}$ \\
$\mathbf{M} \gets {\small \text{HomographyMatrix}(p_s,p_t))}$ \\
$d \gets \text{sqrt}(\text{Trace}( (\mathbf{M}-\mathbf{I})^\intercal*(\mathbf{M}-\mathbf{I})))$ \\
\If{$d < d_{min}$}{
$I_s^*, I_t^* \gets I_s, I_t$
}
}
}
\Return $I_s^*, I_t^*$
\end{algorithm}
\subsection{Labeling and Pairing for UGV Dataset }
Labeling the UGV dataset presents unique challenges due to the difficulty of determining object bounding boxes in hazy images to a human labeler. To address these issues, a 3D object labeling scheme was developed to assist with the labeling process. Initial bounding boxes were generated by projecting the 3D bounding volumes of each object visible in each image, given the UGV's known observation point. To determine the UGV's global position and orientation for each captured image, the Monte-Carlo particle filter-based localization package \emph{AMCL}~\cite{ros} from ROS~\cite{quigley2009ros} was used to localize into a 2D map which was generated for each obstacle configuration using OmniMapper~\cite{trevor2014omnimapper}. The lidar data from the Ouster OS1-64 was used with the UGV's odometry and internal IMU to generate this map. The projections of the 3D object bounding volumes were then refined by hand with LabelMe~\cite{labelme}.
Two UGV trajectories were collected for each object configuration: one with the haze effect and the other without the haze effect. A 2D map was generated for each obstacle configuration and used to localize the UGV with Monte-Carlo localization as described in Section~\ref{sec:ugv_dataset_generation}. This global localization was used to tag each image with the UGV's position and orientation at the capture time. This position and orientation information was used with a Hungarian algorithm to find the optimal pairing between haze and haze-free images, which minimizes the error metric $E=10\Delta_o + \Delta_p$ where $\Delta_o$ is the difference in orientation and $\Delta_p$ is the difference in position. This metric weighs orientation more heavily as changes in orientation have a more significant effect on image viewpoint.
\begin{figure*}[!htb]
\captionsetup[subfigure]{labelformat=empty,justification=centering,farskip=1pt,captionskip=1pt}
\centering
\begin{mdframed}[style=RedBox,nobreak=true]
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/GT/009196.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/GT/018660.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/GT/020100.png}
}
\end{mdframed}
\begin{mdframed}[style=BlueBox,nobreak=true]
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/hazy/009196.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/hazy/018660.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/hazy/020100.png}
}
\end{mdframed}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/FFA-Net/009196.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/FFA-Net/018660.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/FFA-Net/020100.png}
}
\\
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/GCANet/009196.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/GCANet/018660.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/GCANet/020100.png}
}
\\
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/MSBDN/009196.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/MSBDN/018660.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/MSBDN/020100.png}
}
\\
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/SRKT/009196.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/SRKT/018660.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/SRKT/020100.png}
}
\\
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/Trident/009196.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/Trident/018660.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/Trident/020100.png}
}
\\
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/DWDehazing/009196.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/DWDehazing/018660.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/DWDehazing/020100.png}
}
\\
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/Cycle-GAN/9690.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/Cycle-GAN/18660.png}
}
\subfloat
{
\includegraphics[width=0.32\linewidth]{figures/dehaze_images/YOLOX/Cycle-GAN/20100.png}
}
\vspace{-0.5em}
\caption{Examples of detected vehicles on hazy images and dehazed images. All detections are produced from YOLOX-M pretrained on a merged set of VisDrone2019-DET and UAVDT-M. The first row in red frame shows ground-truth bounding-box annotations. The second row in blue frame shows baseline results w/o dehazing. The detected vehicles using different dehazing approaches are arranged from the third row to the last row: FFA-Net, GCANet, MSBDN, SRKT, Trident, DWDehazing, and our proposed Cycle-DehazeNet.
}
\label{fig:detection-benchmarks}
\end{figure*}
\begin{figure}
\centering
\includegraphics[width=1\linewidth]{figures/failed_example/fail_example.pdf}
\caption{Failed examples on dehazed images (detected by YOLOX). YOLOX always mistakenly detect small contrast boards as vehicles (marked in orange on the middle right).}
\label{fig:failed}
\end{figure}
\begin{table*}\normalsize
\centering
\caption{Vehicle Detection on A2I2-Haze w/o Dehazing. The first row of A2I2-UAV shows experimental results on A2I2-UAV w/o NDFT. The second row of A2I2-UAV$^+$ shows experimental results on A2I2-UAV w/ NDFT. The third row of A2I2-UGV shows experimental results on A2I2-UGV.}
\resizebox{0.75\linewidth}{!}{
\begin{tabular}{c|ccc|ccc|ccc|ccc}
\hline
& \multicolumn{3}{c|}{YOLOv5} & \multicolumn{3}{c|}{FasterRCNN} & \multicolumn{3}{c|}{CenterNet} & \multicolumn{3}{c}{YOLOX}\\
\hline
& AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ \\
\hline
A2I2-UAV & 44.9 & 53.8 & 42.8 & 41.1 & 55.3 & 38.1 & 60.9 & 69.3 & 51.4 & 41.2 & 49.9 & 39.5\\
\hline
A2I2-UAV$^+$ & 46.5 & 56.1 & 44.9 & 43.6 & 57.6 & 40.7 & 63.5 & 71.9 & 53.8 & 44.1 & 51.6 & 41.8\\
\hline
A2I2-UGV & 36.5 & 45.5 & 34.2 & 47.8 & 58.0 & 44.3 & 52.9 & 69.3 & 47.9 & 24.0 & 28.7 & 20.9 \\
\hline
\end{tabular}
}
\label{tab:detection_nodehaze}
\end{table*}
\begin{table*}\normalsize
\centering
\caption{Vehicle Detection on A2I2-Haze w/ Dehazing. The upper and lower table shows experiments on A2I2-UAV and A2I2-UGV, respectively. The numbers in black shows the detection score (AR, AP$_{0.5}$, and AP$_{0.5:0.95}$) after applying dehazing on both training and testing images. The numbers in blue shows the relative increase over the detection w/o dehazing. The numbers in red show the relative decrease under the detection w/o dehazing.}
\resizebox{\linewidth}{!}{
\begin{tabular}{c|ccc|ccc|ccc|ccc}
\hline
\multirow{2}*{A2I2-UAV} & \multicolumn{3}{c|}{YOLOv5} & \multicolumn{3}{c|}{FasterRCNN} & \multicolumn{3}{c|}{CenterNet} & \multicolumn{3}{c}{YOLOX}\\
\cline{2-13}
~ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ \\
\hline
GCANet & 50.9(+\blue{6.0}) & 59.7(+\blue{5.9}) & 49.0(+\blue{6.2}) & 47.3(+\blue{6.2}) & 61.1(+\blue{5.8}) & 43.5(+\blue{5.4}) & 65.3(+\blue{4.4}) & 75.2(+\blue{5.9}) & 55.7(+\blue{4.3}) & 40.5(-\red{0.7}) & 47.9(-\red{2.0}) & 38.5(-\red{1.0})\\
FFA-Net & 45.3(+\blue{0.4}) & 53.3(-\red{0.5}) & 42.9(+\blue{0.1}) & 44(+\blue{2.9}) & 54.4(-\red{0.9}) & 37.4(-\red{0.7}) & 61.8(+\blue{0.9}) & 69.6(+\blue{0.3}) & 51.5(+\blue{0.1}) & 40.6(-\red{0.6}) & 48.5(-\red{1.4}) & 38.6(-\red{0.9}) \\
MSBDN & 49.5(+\blue{4.6}) & 58.8(+\blue{5.0}) & 47.5(+\blue{4.7}) & 47.3(+\blue{6.2}) & 62.8(+\blue{7.5}) & 43.6(+\blue{5.5}) & 64.2(+\blue{3.3}) & 72.5(+\blue{3.2}) & 57.5(+\blue{6.1}) & 42.7(+\blue{1.5}) & 48.8(-\red{1.1}) & 39.0(-\red{0.5}) \\
Trident & 56.2(+\blue{11.3}) & 67.3(+\blue{13.5}) & 53.4(+\blue{10.6}) & 55.1(+\blue{14.0}) & 71.6(+\blue{16.3}) & 49.7(+\blue{11.6}) & 65.5(+\blue{4.6}) & 77.3(+\blue{8.0}) & 57.0(+\blue{5.6}) & 40.1(-\red{1.1}) & 48.3(-\red{1.6}) & 38.4(-\red{1.1}) \\
SRKT & 52.4(+\blue{7.5}) & 63.5(+\blue{9.7}) & 49.9(+\blue{7.1}) & 50.7(+\blue{9.6}) & 65.7(+\blue{10.4}) & 45.8(+\blue{7.7}) & 62.9(+\blue{2.0}) & 74.1(+\blue{4.8}) & 54.2(+\blue{2.8}) & 39.4(-\red{1.8}) & 45.9(-\red{4.0}) & 37.0(-\red{2.5}) \\
DWDehaze & 54.6(+\blue{9.7}) & 65.5(+\blue{11.7}) & 52.0(+\blue{9.2}) & 53.1(+\blue{12.0}) & 67.7(+\blue{12.4}) & 48.4(+\blue{10.3}) & 65.5(+\blue{4.6}) & 76.3(+\blue{7.0}) & 56.5(+\blue{5.1}) & \textbf{44.1(+\blue{2.9})} & \textbf{51.6(+\blue{1.7})} & 41.8(+\blue{2.3}) \\
Cycle & \textbf{57.1(+\blue{12.2})} & \textbf{68.1(+\blue{14.3})} & \textbf{53.9(+\blue{11.1})} & \textbf{56.6(+\blue{15.5})} & \textbf{73.1(+\blue{17.8})} & \textbf{50.1(+\blue{12.0})} & \textbf{66.2(+\blue{5.3})} & \textbf{77.8(+\blue{8.5})} & \textbf{57.7(+\blue{6.3})} & 43.9(+\blue{2.7}) & 51.5(+\blue{1.6}) & \textbf{41.8(+\blue{2.3})} \\
\hline
\hline
\multirow{2}*{A2I2-UGV} & \multicolumn{3}{c|}{YOLOv5} & \multicolumn{3}{c|}{FasterRCNN} & \multicolumn{3}{c|}{CenterNet} & \multicolumn{3}{c}{YOLOX}\\
\cline{2-13}
~ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ & AR & $\text{AP}_{0.5}$ & $\text{AP}_{0.5:0.95}$ \\
\hline
GCANet & 38.2(+\blue{1.7}) & 47.5(+\blue{2.0}) & 36.0(+\blue{1.8}) & 54.5(+\blue{6.7}) & 67.6(+\blue{9.6}) & 50.0(+\blue{5.7}) & 54.1(+\blue{1.2}) & \textbf{73.0(+\blue{3.7})} & 49.3(+\blue{1.4}) & 30.2(+\blue{6.2}) & 33.7(+\blue{5.0}) & 24.7(+\blue{3.8})\\
FFA-Net & 37.3(+\blue{0.8}) & 46.5(+\blue{1.0}) & 35.0(+\blue{0.8}) & 48.1(+\blue{0.3}) & 58.0(-\red{0.0}) & 44.3(-\red{0.0}) & 52.1(-\red{0.8}) & 68.5(-\red{0.8}) & 47.1(-\red{0.8}) & 24.1(+\blue{0.1}) & 28.6(-\red{0.1}) & 20.8(-\red{0.1}) \\
MSBDN & 37.1(+\blue{0.6}) & 45.5(-\red{0.0}) & 34.7(+\blue{0.5}) & 50.1(+\blue{2.3}) & 60.9(+\blue{2.9}) & 46.4(+\blue{2.1}) & 53.3(+\blue{0.4}) & 69.7(+\blue{0.4}) & 48.4(+\blue{0.5}) & 32.0(+\blue{8.0}) & 36.2(+\blue{7.5}) & 27.8(+\blue{6.9}) \\
Trident & \textbf{42.4(+\blue{5.9})} & \textbf{53.4(+\blue{7.9})} & \textbf{40.2(+\blue{6.0})} & 53.8(+\blue{6.0}) & 66.5(+\blue{8.5}) & 49.5(+\blue{5.2}) & \textbf{54.2(+\blue{1.3})} & 72.7(+\blue{3.4}) & 48.8(+\blue{0.9}) & 35.0(+\blue{11.0}) & 42.7(+\blue{14.0}) & 31.1(+\blue{10.2}) \\
SRKT & 39.9(+\blue{3.4}) & 49.5(+\blue{4.0}) & 36.9(+\blue{2.7}) & 53.6(+\blue{5.8}) & 65.8(+\blue{7.8}) & 49.0(+\blue{4.7}) & 53.4(+\blue{0.5}) & 71.6(+\blue{2.3}) & 48.7(+\blue{0.8}) & 35.7(+\blue{11.7}) & 43.0(+\blue{14.3}) & 31.4(+\blue{10.4}) \\
DWDehaze & 36.4(-\red{0.1}) & 45.5(-\red{0.0}) & 34.2(-\red{0.0}) & 50.6(+\blue{2.8}) & 62.0(+\blue{4.0}) & 46.6(+\blue{2.3}) & 51.8(-\red{1.1}) & 67.9(-\red{1.4})& 45.9(-\red{2.0}) & 30.0(+\blue{6.0}) & 38.3(+\blue{9.6}) & 26.4(+\blue{5.5}) \\
Cycle & 40.6(+\blue{4.1}) & 50.0(+\blue{4.5}) & 37.1(+\blue{2.9}) & \textbf{55.9(+\blue{8.1})} & \textbf{68.7(+\blue{10.7})} & \textbf{50.5(+\blue{6.2})} & 54.1(+\blue{1.2}) & 72.9(+\blue{3.6}) & \textbf{49.3(+\blue{1.4})} & \textbf{36.2(+\blue{12.2})} & \textbf{44.1(+\blue{15.4})} & \textbf{32.2(+\blue{11.3})} \\
\hline
\end{tabular}
}
\label{tab:A2I2-Haze_Dehaze}
\end{table*}
\begin{figure*}[!htb]
\captionsetup[subfigure]{labelformat=empty,justification=centering,farskip=0pt,captionskip=1pt}
\centering
\subfloat
{
\includegraphics[width=0.475\linewidth]{figures/UAV.pdf}
}
\subfloat{
\includegraphics[width=0.475\linewidth]{figures/UGV.pdf}
}
\caption{Comparison of different detection and dehazing approaches. The left plot shows detection results on A2I2-UAV, and the right plot shows detection results on A2I2-UGV. Scatter points in these two plots correspond to AP$_{0.5:0.95}$ in Table~\ref{tab:A2I2-Haze_Dehaze}. It gives two perspective: object detectors (the red line) and dehazing approaches (the gray lines). In the left plot, the dashed gray line with triangles is clearly above the remaining three gray lines, which indicates that CenterNet is the best detector in non-homogeneous haze. In the dashed red line with stars, our proposed Cycle-DehazeNet has the best dehazing performance, indicated by the largest improvement on the four detectors' performance.}
\label{fig:detection_dehaze}
\end{figure*}
\begin{figure}[!htb]
\captionsetup[subfigure]{labelformat=empty,justification=centering,farskip=1pt,captionskip=1pt}
\centering
\subfloat
{
\includegraphics[width=1.0\linewidth]{figures/cycle-dehaze/Cycle-dehaze-New.pdf}
}
\caption{Overview of Cycle-DehazeNet. $\rm G_{H2C}$ and $\rm G_{C2H}$ are generators, while $\rm D_{clean}$ and $\rm D_{hazy}$ are discriminators. Note that $\rm G_{H2C}$ translates image from hazy domain to haze-free domain while $\rm G_{C2H}$ translates image from haze-free domain to hazy domain. $\rm D_{clean}$ distinguishes real haze-free images from fake ones. $\rm D_{hazy}$ tells real hazy images from fake ones. VGG is used to give feature-level perceptual loss supervision.}
\label{fig:cycle-dehaze}
\end{figure}
\section{Quantitative Evaluation}
Within A2I2-Haze dataset, we created two separate subsets, A2I2-UAV and A2I2-UGV, dedicated to UAV and UGV sets, respectively.
A2I2-UAV includes a training set UAV-train with $224$ pairs of hazy and corresponding haze-free images, and an additional $240$ haze-free images. The testing set of UAV-test has $119$ hazy images. The training set of A2I2-UGV - UGV-train has $50$ pairs of hazy and corresponding haze-free images, and an additional $200$ haze-free images. UGV-test has $200$ hazy images.
\subsection{Object Detection}\label{det_eval}
\subsubsection{Detection Approaches}
We use four state-of-the-art detectors that are widely used both in industry and academia: (a) \textit{YOLOv5}~\cite{yolov5}, (b) \textit{YOLOX}~\cite{ge2021yolox}, (c) \textit{Faster R-CNN}~\cite{ren2016faster}, and (d) \textit{CenterNet}~\cite{zhou2019objects} to evaluate the proposed A2I2-Haze dataset. For experiments on A2I2-UAV, all detectors are pretrained on a merged set of VisDrone2019-DET~\cite{cao2021visdrone} and UAVDT-M~\cite{du2018unmanned} and adopt official COCO-based implementation. For experiments on A2I2-UGV, all detectors are pretrained on Cityscapes~\cite{cordts2016cityscapes} and adopt official COCO-based implementation.
\subsubsection{Results and Analyses} Table~\ref{tab:detection_nodehaze} shows the performance of various object detectors without dehazing on both A2I2-UAV and A2I2-UGV subsets of A2I2-Haze. In the first row, we infer that object detectors can be ordered by their detection performance on A2I2-UAV as follows: $\text{CenterNet} > \text{YOLOv5} > \text{YOLOX} > \text{FasterRCNN}$. In the second row, we show that utilizing the collected altitude metadata to learn an altitude variation robust detector by NDFT~\cite{wu2019delving} could improve the detection performance for all the four detectors. For A2I2-UGV (the third row), the detectors can be ordered based on their performance as $\text{CenterNet} > \text{FasterRCNN} > \text{YOLOv5} > \text{YOLOX}$. Among the four detectors, YOLOX consistently gives the worst performance on both A2I2-UAV and A2I2-UGV. It frequently mistakes small contrast boards as vehicles, as shown in Fig.~\ref{fig:failed}.
\subsection{Dehazing}
\subsubsection{Baseline Approaches}
We benchmark three state-of-the-art homogeneous dehazing approaches: (I) \textit{GCANet}~\cite{chen2019gated}, (II) \textit{FFA-Net}~\cite{qin2020ffa}, and (III) \textit{MSBDN}~\cite{dong2020multi}. In addition, we also benchmark another set of three state-of-the-art non-homogeneous dehazing approaches: (IV) \textit{SRKT}~\cite{chen2021srktdn}, (V) \textit{DWDehaze}~\cite{fu2021dw}, and (VI) \textit{Trident-Dehazing}~\cite{liu2020trident}. All dehazing models are pre-trained using official implementations.
\subsubsection{Proposed Cycle-DehazeNet}
Inspired by CycleGAN~\cite{engin2018cycle,you2019ct,zhu2017unpaired}, we propose a dehazing model that could be trained for image translation tasks on the hazy image domain and haze-free image domain without paired training samples.
%
Cycle-DehazeNet is trained on $512\times 512$ patches (CycleGAN's constraint). It takes $512 \times 512$ as input resolution and restores $512 \times 512$ output images to minimize computational costs. In the inference, we first pad the input HR image from $1845 \times 1500$ to $2048 \times 1536$, so that the width and height are divisible by $512$. Then we extract overlapping $512 \times 512$ patches at a stride of $256$. Each target pixel $p_t$ in the image may be covered by multiple patches. Last, the target pixel value is computed as a weighted sum of all the collided pixels $p_i$ from different patches, where the weight is measured by the distance between the pixel and the patch center. The weight is decayed by a Gaussian distribution with controllable sigma parameters.
It could be mathematically expressed below:
\begin{equation}
p_t = \frac{\sum_{i=1}^N w_i p_i}{\sum_{i=1}^N w_i},
\end{equation}
where $N$ is the number of patches that cover the subject pixel and $w$ follows a Gaussian distribution w.r.t. the geometric distance between the patch center and the target pixel.
%
Following \cite{engin2018cycle}, Cycle-DehazeNet has two losses: cycle-consistency loss~\cite{zhu2017unpaired} and perceptual loss~\cite{johnson2016perceptual}.
%
Cycle-consistency loss relieves the constraint of paired training samples. Perceptual loss remedies the local texture information that is heavily corrupted by haze and preserves the original image structure by analyzing the high and low-level features extracted from VGG Net.
%
% We observe that the hazy area is typically further away from the image center in A2I2-Haze.
%
We observe that the hazy areas in our dataset are mostly distributed around four corners instead of the central part. Based on that, we would like to force our proposed method to more \textbf{focus} on haze rather than clean parts. Thus the cropped four $512 \times 512$ patches from the corners of each image are used as ``focus'' (red bounding boxes in Fig.~\ref{fig:cycle-dehaze}). We emphasize more on the hazy area than the whole image by assigning larger weights for the ``focus'' patch in the final loss.
\subsubsection{Evaluation Metric}
We use object detection score on the dehazed images as the metric to quantitatively measure restoration capability of the dehazing technique on image semantics. We use the same four detectors as in Section.~\ref{det_eval} (a) \textit{YOLOv5}, (b) \textit{YOLOX}, (c) \textit{Faster R-CNN}, and (d) \textit{CenterNet}.
\subsubsection{Results and Analyses}
Fig.~\ref{fig:detection-benchmarks} shows detection results on original hazy images and dehazed images. Fig.~\ref{fig:detection_dehaze} shows AP$_{0.5:0.95}$ score as metric to compare various detectors and various dehazing approaches. Fig.~\ref{fig:detection_dehaze} gives two perspective: object detectors (the red line) and dehazing approaches (the gray lines). The red line shows the averaged AP$_{0.5:0.95}$ across the four detectors for different dehazing approaches. The gray lines show the AP$_{0.5:0.95}$ across different dehazing approaches for the four detectors.
Fig.~\ref{fig:detection-benchmarks} shows the visual effect of various dehazing approaches on three randomly selected images. Among the seven dehazing approaches used, FFA-Net leads to the worst detection performance, and our proposed Cycle-DehazeNet gives the best mAP on YOLOv5, Faster R-CNN and CenterNet. In general, we observe that non-homogeneous dehazing algorithms have better mAP compared to the homogeneous dehazing algorithms.
\section{Discussions and Future Works}
A2I2-Haze being the first attempt to develop a real-world hazy dataset with in-situ measurements, we acknowledge several limitations in the data collection procedure. We propose the following research directions to address the gaps:
\begin{itemize}[leftmargin=*]
\item \textbf{Object Diversity:} Both UAV and UGV datasets have a limited number of object categories and instances. Improving the diversity in objects will be a crucial step towards developing a more comprehensive dataset.
\item \textbf{Background Diversity:} A2I2-Haze was primarily collected in two different scenarios: with concrete and grass backgrounds. Including additional scenes with diverse backgrounds will make the dataset a more challenging benchmark.
\item \textbf{Global Haze Measure:} The smoke measurement for A2I2-Haze was made using two laser transmissometers. These measurements can be highly localized in a non-homogeneous setting. A novel sensor network for grid-based measurement is required to improve the accuracy of the in-situ data.
\item \textbf{Synchronized Aerial-Ground View:} A2I2-Haze could be extended to a multi-view dataset using air-ground coordination and multi-source image matching.
\item \textbf{Jointly Optimized Dehazing and Detection Pipeline:}
If jointly optimized, low-level image processing could be made beneficial for high-level semantic tasks~\cite{liu2020connecting,li2017aod}. In this study, dehazing can be considered a pre-processing step for the subsequent detection task. Thus, as a future research direction, detection and dehazing could be jointly designed and trained to optimize detection performance in the presence of haze. This could be particularly significant in the case of aerial-view imagery. Furthermore, as this approach removes the constraint of maintaining the aesthetic quality of the image, dehazing could be utilized to purely restore image semantics for object detection.
\item \textbf{Metadata Utilization:}
The annotated UAV-specific nuisance or metadata in A2I2-Haze could be utilized to improve the robustness of the learned features via adversarial training as discussed in NDFT~\cite{wu2019delving}. In Tab.~\ref{tab:detection_nodehaze}, we have shown some preliminary success by learning a detector robust to altitude variation on A2I2-UAV. Utilizing the haze density metadata to train a dehazing-aware detector better could be a future research direction.
\end{itemize}
\section{Conclusion }
In this paper we develop a challenging UAV and UGV dataset with realistic haze for foundational research in scene understanding in obscured conditions. Advanced smoke-generation and measurement techniques were used to develop this dataset for object detection and dehazing tasks. All images were curated precisely using a combination of both manual and partially-automated techniques. Further, the images are synchronized with haze density measurements and altitude of the UAV. Quantitative evaluation was performed using state-of-the-art object detectors and de-hazing models. Overall the baseline approaches were found to perform poorly on A2I2-Haze and we hope the dataset will serve as a good benchmark for future algorithmic advances. As discussed in the paper, we also plan to address the limitations of A2I2-Haze in future research efforts.
% biography section
%
% If you have an EPS/PDF photo (graphicx package needed) extra braces are
% needed around the contents of the optional argument to biography to prevent
% the LaTeX parser from getting confused when it sees the complicated
% \includegraphics command within an optional argument. (You could create
% your own custom macro containing the \includegraphics command to make things
% simpler here.)
%\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{mshell}}]{Michael Shell}
% or if you just want to reserve a space for a photo:
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/pn.jpg}}]{Priya Narayanan} received her Ph.D. from the Mechanical Engineering Department at University of Maryland Baltimore County (UMBC). She was a National Research Council (NRC) Fellow in the Navy Center for Applied Research in Artificial Intelligence (NCARAI) at the Naval Research Laboratory (NRL). Currently she is a Engineering Researcher at DEVCOM Army Research Laboratory. She has led key research programs in aerial and ground robotics including perception integration activities for the final capstone experimentation of Robotic Collaborative Technology Alliance (Robotic CTA) Program.
\end{IEEEbiography}
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/Xin.jpg}}]{Xin Hu}
received the B.Eng. from Huazhong University of Science and Technology in 2019 and M.S.
from Texas A\&M University in 2021. He is currently a Ph.D. student advised by Dr. Zhengming (Allan) Ding at Tulane University. His research interests include decision-making in autonomous vehicles, video understanding, and multi-modality.
\end{IEEEbiography}
% if you will not have a photo at all:
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/ZhenyuWu.jpg}}]{Zhenyu Wu}
received the B.Eng. from Shanghai Jiao Tong University in 2015 and Ph.D. from Texas A\&M University in 2021. His Ph.D. advisor is Dr. Zhangyang Wang. His research interests include object detection, video understanding, efficient vision on embedded systems, and adversarial machine learning.
He is currently a researcher at Wormpex AI Research. He is closely working with Dr. Zhou Ren, Dr. Yi Wu and Dr. Gang Hua.
\end{IEEEbiography}
% insert where needed to balance the two columns on the last page with
% biographies
%\newpage
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/MThielke-photo.jpg}}] {Matthew Thielke}
received his B.S. in Physics from Virginia Tech. He worked at the DEVCOM C5ISR’s Night Vision and Electronic Sensors Directorate in Electro-Optic measurements and processing. Currently at DEVCOM U.S. Army Research Laboratory, he has extensive measurements experience with Electro-Optical sensors including infrared imagers, and hyper spectral imagers. He has led data collection efforts in thermal and visible face recognition including a ``Large-Scale, Time-Synchronized Visible and Thermal Face Dataset.''
\end{IEEEbiography}
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/John_Rogers.jpg}}] {John G Rogers}
received his Ph.D. in Robotics at the Georgia Institute of Technology, his masters in Computer Science at Stanford, and his masters and bachelors degrees in Electrical and Computer Engineering at Carnegie Mellon. He is currently a senior research scientist at the US DEVCOM Army Research Laboratory (ARL). He is currently leading research efforts in coordinated tactical maneuver in complex and contested terrain for teams of ground robots at ARL.
\end{IEEEbiography}
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/AndreHarrison_sp19.jpg}}]{Andre V. Harrison}
received his Ph.D. in Electrical and Computer Engineering from Johns Hopkins University, his M.E. and B.E. in Electrical and Computer Engineering from Cornell University. He is currently a Computer Engineer with the US DEVCOM Army Research Laboratory (ARL). He has led key research efforts in visual perception for ground vehicles as part of the A.I. for Maneuver and Mobility Essential Research Program. He has also led research projects estimating visual salience and predicting user state.
\end{IEEEbiography}
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/john_dagostino.jpg}}] {John D'Agostino}
received his B.S. from The Pennsylvania State University in Mechanical Engineering. He is currently a Mechanical Engineer with the U.S Army DEVCOM Chemical and Biological Center in the Research and Technology Directorate. He leads a variety of countermeasure activities including the testing of sensors in various degraded battlefield conditions.
\end{IEEEbiography}
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/jamesbrown.jpg}}] {James D Brown}
is a Mechanical Engineer and contractor supporting the US Army DEVCOM-CBC Advanced Design and Manufacturing division. He has supported a variety of UAV and UGV projects from design to implementation, and served as the primary ground control station operator during DVE testing.
\end{IEEEbiography}
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/Long_Quang.png}}] {Long Quang}
acquired a Bachelor of Science in Electrical Engineering at the University of Texas at Dallas. He is currently an Electronics Engineer with the U.S. Army DEVCOM Computational and Information Sciences Directorate, where he actively facilitates robotics research and systems integration.
\end{IEEEbiography}
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/Uplinger.jpg}}] {James Uplinger}
received his doctorate and masters from the Physics
Department at the University of Arkansas, and bachelors from the Physics
Department at the Rochester Institute of Technology (RIT). Currently he is
an Associate Principal Data Scientist at Huntington Ingalls Industries, supporting DEVCOM Army Research Laboratory. He has led
research programs in synthetic infrared (IR) image generation, Radio
Frequency (RF) side channel analytics, and biotechnology development.
\end{IEEEbiography}
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/HeesungKwon_new.jpg}}] {Heesung Kwon}
is Senior Researcher and Team Lead at the DEVCOM Army Research Laboratory (ARL). He received the B.Sc. degree in Electronic Engineering from Sogang University, Seoul, Korea, in 1984, and the MS and Ph.D. degrees in Electrical Engineering from the State University of New York at Buffalo in 1995 and 1999, respectively. Dr. Kwon rejoined ARL in August 2007 as Team Lead and has been leading various AI/ML efforts pertaining to semantic scene understanding in resource-constrained environments, unmanned aerial systems (UAS)-based perception and action/activity recognition at the edge, etc., primarily leveraging machine learning-based approaches.
\end{IEEEbiography}
\begin{IEEEbiography}[{\includegraphics[width=1in,height=1.25in,clip,keepaspectratio]{figures/bio_images/atlaswang.jpeg}}]{Zhangyang Wang}
is currently an Assistant Professor of ECE at UT Austin. He received his Ph.D. in ECE from UIUC in 2016, and his B.E. in EEIS from USTC in 2012. Prof. Wang is broadly interested in the fields of machine learning, computer vision, optimization, and their interdisciplinary applications. His latest interests focus on automated machine learning (AutoML), learning-based optimization, machine learning robustness, and efficient deep learning.
\end{IEEEbiography}
% You can push biographies down or up by placing
% a \vfill before or after them. The appropriate
% use of \vfill depends on what kind of text is
% on the last page and whether or not the columns
% are being equalized.
%\vfill
% Can be used to pull up biographies so that the bottom of the last one
% is flush with the other column.
%\enlargethispage{-5in}
% Generated by IEEEtran.bst, version: 1.12 (2007/01/11)
\begin{thebibliography}{10}
\providecommand{\url}[1]{#1}
\csname url@samestyle\endcsname
\providecommand{\newblock}{\relax}
\providecommand{\bibinfo}[2]{#2}
\providecommand{\BIBentrySTDinterwordspacing}{\spaceskip=0pt\relax}
\providecommand{\BIBentryALTinterwordstretchfactor}{4}
\providecommand{\BIBentryALTinterwordspacing}{\spaceskip=\fontdimen2\font plus
\BIBentryALTinterwordstretchfactor\fontdimen3\font minus
\fontdimen4\font\relax}
\providecommand{\BIBforeignlanguage}[2]{{%
\expandafter\ifx\csname l@#1\endcsname\relax
\typeout{** WARNING: IEEEtran.bst: No hyphenation pattern has been}%
\typeout{** loaded for the language `#1'. Using the pattern for}%
\typeout{** the default language instead.}%
\else
\language=\csname l@#1\endcsname
\fi
#2}}
\providecommand{\BIBdecl}{\relax}
\BIBdecl
\bibitem{li2019benchmarking}
B.~Li, W.~Ren, D.~Fu, D.~Tao, D.~Feng, W.~Zeng, and Z.~Wang, ``Benchmarking
single-image dehazing and beyond,'' \emph{IEEE Transactions on Image
Processing}, vol.~28, no.~1, pp. 492--505, 2019.
\bibitem{NH-Haze_2020}
C.~O. Ancuti, C.~Ancuti, and R.~Timofte, ``{NH-HAZE:} an image dehazing
benchmark with non-homogeneous hazy and haze-free images,'' in
\emph{Proceedings of the IEEE Conference on Computer Vision and Pattern
Recognition Workshops}, ser. IEEE CVPR 2020, 2020.
\bibitem{REVIDE}
X.~Zhang, H.~Dong, J.~Pan, C.~Zhu, Y.~Tai, C.~Wang, J.~Li, F.~Huang, and
F.~Wang, ``Learning to restore hazy video: A new real-world dataset and a new
method,'' in \emph{CVPR}, 2021, pp. 9239--9248.
\bibitem{zhang2020nighttime}
J.~Zhang, Y.~Cao, Z.-J. Zha, and D.~Tao, ``Nighttime dehazing with a synthetic
benchmark,'' in \emph{Proceedings of the 28th ACM International Conference on
Multimedia}, 2020, pp. 2355--2363.
\bibitem{Zhang:HazeRD:ICIP17b}
\BIBentryALTinterwordspacing
Y.~Zhang, L.~Ding, and G.~Sharma, ``Hazerd: an outdoor scene dataset and
benchmark for single image dehazing,'' in \emph{Proc. IEEE Intl. Conf. Image
Proc.}, 2017, pp. 3205--3209. [Online]. Available:
\url{http://www.ece.rochester.edu/~gsharma/papers/Zhang_ICIP2017_HazeRD.pdf,
paper
https://labsites.rochester.edu/gsharma/research/computer-vision/hazerd/,
project page and dataset}
\BIBentrySTDinterwordspacing
\bibitem{4K_Xiao_2022}
\BIBentryALTinterwordspacing
B.~Xiao, Z.~Zheng, X.~Chen, C.~Lv, Y.~Zhuang, and T.~Wang, ``Single uhd image
dehazing via interpretable pyramid network,'' 2022. [Online]. Available:
\url{https://arxiv.org/abs/2202.08589}
\BIBentrySTDinterwordspacing
\bibitem{liu2020end}
W.~Liu, X.~Hou, J.~Duan, and G.~Qiu, ``End-to-end single image fog removal
using enhanced cycle consistent adversarial networks,'' \emph{IEEE
Transactions on Image Processing}, vol.~29, pp. 7819--7833, 2020.
\bibitem{zhao2020dehazing}
S.~Zhao, L.~Zhang, S.~Huang, Y.~Shen, and S.~Zhao, ``Dehazing evaluation:
Real-world benchmark datasets, criteria, and baselines,'' \emph{IEEE
Transactions on Image Processing}, vol.~29, pp. 6947--6962, 2020.
\bibitem{ancuti2019dense}
C.~O. Ancuti, C.~Ancuti, M.~Sbert, and R.~Timofte, ``Dense-haze: A benchmark
for image dehazing with dense-haze and haze-free images,'' in \emph{2019 IEEE
international conference on image processing (ICIP)}.\hskip 1em plus 0.5em
minus 0.4em\relax IEEE, 2019, pp. 1014--1018.
\bibitem{O-HAZE_2018}
C.~O. Ancuti, C.~Ancuti, R.~Timofte, and C.~D. Vleeschouwer, ``O-haze: a
dehazing benchmark with real hazy and haze-free outdoor images,'' in
\emph{IEEE Conference on Computer Vision and Pattern Recognition, NTIRE
Workshop}, ser. NTIRE CVPR'18, 2018.
\bibitem{I-HAZE_2018}
------, ``I-haze: a dehazing benchmark with real hazy and haze-free indoor
images,'' in \emph{arXiv:1804.05091v1}, 2018.
\bibitem{mccartney1976optics}
E.~J. McCartney, ``Optics of the atmosphere: scattering by molecules and
particles,'' \emph{New York}, 1976.
\bibitem{nayar1999vision}
S.~K. Nayar and S.~G. Narasimhan, ``Vision in bad weather,'' in
\emph{Proceedings of the seventh IEEE international conference on computer
vision}, vol.~2.\hskip 1em plus 0.5em minus 0.4em\relax IEEE, 1999, pp.
820--827.
\bibitem{narasimhan2003contrast}
S.~G. Narasimhan and S.~K. Nayar, ``Contrast restoration of weather degraded
images,'' \emph{IEEE transactions on pattern analysis and machine
intelligence}, vol.~25, no.~6, pp. 713--724, 2003.
\bibitem{peng2007image}
H.~Peng and R.~Rao, ``Image enhancement of fog-impaired scenes with variable
visibility,'' in \emph{2007 IEEE International Conference on Acoustics,
Speech and Signal Processing-ICASSP'07}, vol.~2.\hskip 1em plus 0.5em minus
0.4em\relax IEEE, 2007, pp. II--389.
\bibitem{rao2005algorithms}
R.~Rao and S.~Lee, ``Algorithms for scene restoration and visibility estimation
from aerosol scatter impaired images,'' in \emph{IEEE International
Conference on Image Processing 2005}, vol.~1.\hskip 1em plus 0.5em minus
0.4em\relax IEEE, 2005, pp. I--929.
\bibitem{berman2016non}
D.~Berman, S.~Avidan \emph{et~al.}, ``Non-local image dehazing,'' in
\emph{Proceedings of the IEEE conference on computer vision and pattern
recognition}, 2016, pp. 1674--1682.
\bibitem{fattal2008single}
R.~Fattal, ``Single image dehazing,'' \emph{ACM transactions on graphics
(TOG)}, vol.~27, no.~3, pp. 1--9, 2008.
\bibitem{fattal2014dehazing}
------, ``Dehazing using color-lines,'' \emph{ACM transactions on graphics
(TOG)}, vol.~34, no.~1, pp. 1--14, 2014.
\bibitem{he2010single}
K.~He, J.~Sun, and X.~Tang, ``Single image haze removal using dark channel
prior,'' \emph{IEEE transactions on pattern analysis and machine
intelligence}, vol.~33, no.~12, pp. 2341--2353, 2010.
\bibitem{tan2008visibility}
R.~T. Tan, ``Visibility in bad weather from a single image,'' in \emph{2008
IEEE conference on computer vision and pattern recognition}.\hskip 1em plus
0.5em minus 0.4em\relax IEEE, 2008, pp. 1--8.
\bibitem{zhu2015fast}
Q.~Zhu, J.~Mai, and L.~Shao, ``A fast single image haze removal algorithm using
color attenuation prior,'' \emph{IEEE transactions on image processing},
vol.~24, no.~11, pp. 3522--3533, 2015.
\bibitem{ren2016single}
W.~Ren, S.~Liu, H.~Zhang, J.~Pan, X.~Cao, and M.-H. Yang, ``Single image
dehazing via multi-scale convolutional neural networks,'' in \emph{European
conference on computer vision}.\hskip 1em plus 0.5em minus 0.4em\relax
Springer, 2016, pp. 154--169.
\bibitem{cai2016dehazenet}
B.~Cai, X.~Xu, K.~Jia, C.~Qing, and D.~Tao, ``Dehazenet: An end-to-end system
for single image haze removal,'' \emph{IEEE Transactions on Image
Processing}, vol.~25, no.~11, pp. 5187--5198, 2016.
\bibitem{zhang2018densely}
H.~Zhang and V.~M. Patel, ``Densely connected pyramid dehazing network,'' in
\emph{Proceedings of the IEEE conference on computer vision and pattern
recognition}, 2018, pp. 3194--3203.
\bibitem{li2018single}
R.~Li, J.~Pan, Z.~Li, and J.~Tang, ``Single image dehazing via conditional
generative adversarial network,'' in \emph{Proceedings of the IEEE Conference
on Computer Vision and Pattern Recognition}, 2018, pp. 8202--8211.
\bibitem{li2017aod}
B.~Li, X.~Peng, Z.~Wang, J.~Xu, and D.~Feng, ``Aod-net: All-in-one dehazing
network,'' in \emph{Proceedings of the IEEE international conference on
computer vision}, 2017, pp. 4770--4778.
\bibitem{li2018end}
------, ``End-to-end united video dehazing and detection,'' in
\emph{Proceedings of the AAAI Conference on Artificial Intelligence},
vol.~32, no.~1, 2018.
\bibitem{ren2018gated}
W.~Ren, L.~Ma, J.~Zhang, J.~Pan, X.~Cao, W.~Liu, and M.-H. Yang, ``Gated fusion
network for single image dehazing,'' in \emph{Proceedings of the IEEE
Conference on Computer Vision and Pattern Recognition}, 2018, pp. 3253--3261.
\bibitem{qu2019enhanced}
Y.~Qu, Y.~Chen, J.~Huang, and Y.~Xie, ``Enhanced pix2pix dehazing network,'' in
\emph{Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern
Recognition}, 2019, pp. 8160--8168.
\bibitem{dong2020multi}
H.~Dong, J.~Pan, L.~Xiang, Z.~Hu, X.~Zhang, F.~Wang, and M.-H. Yang,
``Multi-scale boosted dehazing network with dense feature fusion,'' in
\emph{Proceedings of the IEEE/CVF conference on computer vision and pattern
recognition}, 2020, pp. 2157--2167.
\bibitem{zheng2021ultra}
Z.~Zheng, W.~Ren, X.~Cao, X.~Hu, T.~Wang, F.~Song, and X.~Jia,
``Ultra-high-definition image dehazing via multi-guided bilateral learning,''
in \emph{2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition
(CVPR)}.\hskip 1em plus 0.5em minus 0.4em\relax IEEE, 2021, pp.
16\,180--16\,189.
\bibitem{shyam2021towards}
P.~Shyam, K.-J. Yoon, and K.-S. Kim, ``Towards domain invariant single image
dehazing,'' in \emph{Proceedings of the AAAI Conference on Artificial
Intelligence}, vol.~35, no.~11, 2021, pp. 9657--9665.
\bibitem{yeh2019multi}
C.-H. Yeh, C.-H. Huang, and L.-W. Kang, ``Multi-scale deep residual
learning-based single image haze removal via image decomposition,''
\emph{IEEE Transactions on Image Processing}, vol.~29, pp. 3153--3167, 2019.
\bibitem{deng2019deep}
Z.~Deng, L.~Zhu, X.~Hu, C.-W. Fu, X.~Xu, Q.~Zhang, J.~Qin, and P.-A. Heng,
``Deep multi-model fusion for single-image dehazing,'' in \emph{Proceedings
of the IEEE/CVF international conference on computer vision}, 2019, pp.
2453--2462.
\bibitem{li2019lap}
Y.~Li, Q.~Miao, W.~Ouyang, Z.~Ma, H.~Fang, C.~Dong, and Y.~Quan, ``Lap-net:
Level-aware progressive network for image dehazing,'' in \emph{Proceedings of
the IEEE/CVF International Conference on Computer Vision}, 2019, pp.
3276--3285.
\bibitem{girshick2014rich}
R.~Girshick, J.~Donahue, T.~Darrell, and J.~Malik, ``Rich feature hierarchies
for accurate object detection and semantic segmentation,'' in
\emph{Proceedings of the IEEE conference on computer vision and pattern
recognition}, 2014, pp. 580--587.
\bibitem{girshick2015fast}
R.~Girshick, ``Fast r-cnn,'' in \emph{Proceedings of the IEEE international
conference on computer vision}, 2015, pp. 1440--1448.
\bibitem{ren2016faster}
S.~Ren, K.~He, R.~Girshick, and J.~Sun, ``Faster r-cnn: Towards real-time
object detection with region proposal networks,'' \emph{TPAMI}, 2016.
\bibitem{dai2016r}
J.~Dai, Y.~Li, K.~He, and J.~Sun, ``R-fcn: Object detection via region-based
fully convolutional networks,'' \emph{Advances in neural information
processing systems}, vol.~29, 2016.
\bibitem{he2017mask}
K.~He, G.~Gkioxari, P.~Doll{\'a}r, and R.~Girshick, ``Mask r-cnn,'' in
\emph{Proceedings of the IEEE international conference on computer vision},
2017, pp. 2961--2969.
\bibitem{redmon2016you}
J.~Redmon, S.~Divvala, R.~Girshick, and A.~Farhadi, ``You only look once:
Unified, real-time object detection,'' in \emph{Proceedings of the IEEE
conference on computer vision and pattern recognition}, 2016, pp. 779--788.
\bibitem{redmon2017yolo9000}
J.~Redmon and A.~Farhadi, ``Yolo9000: better, faster, stronger,'' in
\emph{Proceedings of the IEEE conference on computer vision and pattern
recognition}, 2017, pp. 7263--7271.
\bibitem{redmon2018yolov3}
------, ``Yolov3: An incremental improvement,'' \emph{arXiv preprint
arXiv:1804.02767}, 2018.
\bibitem{bochkovskiy2020yolov4}
A.~Bochkovskiy, C.-Y. Wang, and H.-Y.~M. Liao, ``Yolov4: Optimal speed and
accuracy of object detection,'' \emph{arXiv preprint arXiv:2004.10934}, 2020.
\bibitem{ge2021yolox}
Z.~Ge, S.~Liu, F.~Wang, Z.~Li, and J.~Sun, ``Yolox: Exceeding yolo series in
2021,'' \emph{arXiv preprint arXiv:2107.08430}, 2021.
\bibitem{liu2016ssd}
W.~Liu, D.~Anguelov, D.~Erhan, C.~Szegedy, S.~Reed, C.-Y. Fu, and A.~C. Berg,
``Ssd: Single shot multibox detector,'' in \emph{European conference on
computer vision}.\hskip 1em plus 0.5em minus 0.4em\relax Springer, 2016, pp.
21--37.
\bibitem{lin2017focal}
T.-Y. Lin, P.~Goyal, R.~Girshick, K.~He, and P.~Doll{\'a}r, ``Focal loss for
dense object detection,'' in \emph{Proceedings of the IEEE international
conference on computer vision}, 2017, pp. 2980--2988.
\bibitem{zhou2019objects}
X.~Zhou, D.~Wang, and P.~Kr{\"a}henb{\"u}hl, ``Objects as points,'' \emph{arXiv
preprint arXiv:1904.07850}, 2019.
\bibitem{tian2019fcos}
Z.~Tian, C.~Shen, H.~Chen, and T.~He, ``Fcos: Fully convolutional one-stage
object detection,'' in \emph{Proceedings of the IEEE/CVF international
conference on computer vision}, 2019, pp. 9627--9636.
\bibitem{he2016deep}
K.~He, X.~Zhang, S.~Ren, and J.~Sun, ``Deep residual learning for image
recognition,'' in \emph{Proceedings of the IEEE conference on computer vision
and pattern recognition}, 2016, pp. 770--778.
\bibitem{howard2017mobilenets}
A.~G. Howard, M.~Zhu, B.~Chen, D.~Kalenichenko, W.~Wang, T.~Weyand,
M.~Andreetto, and H.~Adam, ``Mobilenets: Efficient convolutional neural
networks for mobile vision applications,'' \emph{arXiv preprint
arXiv:1704.04861}, 2017.
\bibitem{lin2017feature}
T.-Y. Lin, P.~Doll{\'a}r, R.~Girshick, K.~He, B.~Hariharan, and S.~Belongie,
``Feature pyramid networks for object detection,'' in \emph{Proceedings of
the IEEE conference on computer vision and pattern recognition}, 2017, pp.
2117--2125.
\bibitem{liu2018path}
S.~Liu, L.~Qi, H.~Qin, J.~Shi, and J.~Jia, ``Path aggregation network for
instance segmentation,'' in \emph{Proceedings of the IEEE conference on
computer vision and pattern recognition}, 2018, pp. 8759--8768.
\bibitem{tan2020efficientdet}
M.~Tan, R.~Pang, and Q.~V. Le, ``Efficientdet: Scalable and efficient object
detection,'' in \emph{Proceedings of the IEEE/CVF conference on computer
vision and pattern recognition}, 2020, pp. 10\,781--10\,790.
\bibitem{ghiasi2019fpn}
G.~Ghiasi, T.-Y. Lin, and Q.~V. Le, ``Nas-fpn: Learning scalable feature
pyramid architecture for object detection,'' in \emph{Proceedings of the
IEEE/CVF Conference on Computer Vision and Pattern Recognition}, 2019, pp.
7036--7045.
\bibitem{carion2020end}
N.~Carion, F.~Massa, G.~Synnaeve, N.~Usunier, A.~Kirillov, and S.~Zagoruyko,
``End-to-end object detection with transformers,'' in \emph{European
conference on computer vision}.\hskip 1em plus 0.5em minus 0.4em\relax
Springer, 2020, pp. 213--229.
\bibitem{ge2021ota}
Z.~Ge, S.~Liu, Z.~Li, O.~Yoshie, and J.~Sun, ``Ota: Optimal transport
assignment for object detection,'' in \emph{Proceedings of the IEEE/CVF
Conference on Computer Vision and Pattern Recognition}, 2021, pp. 303--312.
\bibitem{trevor2014omnimapper}
A.~J. Trevor, J.~G. Rogers, and H.~I. Christensen, ``Omnimapper: A modular
multimodal mapping framework,'' in \emph{2014 IEEE international conference
on robotics and automation (ICRA)}.\hskip 1em plus 0.5em minus 0.4em\relax
IEEE, 2014, pp. 1983--1990.
\bibitem{wu2019delving}
Z.~Wu, K.~Suresh, P.~Narayanan, H.~Xu, H.~Kwon, and Z.~Wang, ``Delving into
robust object detection from unmanned aerial vehicles: A deep nuisance
disentanglement approach,'' in \emph{Proceedings of the IEEE/CVF
International Conference on Computer Vision}, 2019, pp. 1201--1210.
\bibitem{sun2021loftr}
J.~Sun, Z.~Shen, Y.~Wang, H.~Bao, and X.~Zhou, ``Loftr: Detector-free local
feature matching with transformers,'' in \emph{Proceedings of the IEEE/CVF
conference on computer vision and pattern recognition}, 2021, pp. 8922--8931.
\bibitem{ros}
``Ros navigation stack: A 2d navigation stack that takes in information from
odometry, sensor streams, and a goal pose and outputs safe velocity commands
that are sent to a mobile base.''
\url{https://github.com/ros-planning/navigation}.
\bibitem{quigley2009ros}
M.~Quigley, K.~Conley, B.~Gerkey, J.~Faust, T.~Foote, J.~Leibs, R.~Wheeler,
A.~Y. Ng \emph{et~al.}, ``Ros: an open-source robot operating system,'' in
\emph{ICRA workshop on open source software}, vol.~3, no. 3.2.\hskip 1em plus
0.5em minus 0.4em\relax Kobe, Japan, 2009, p.~5.
\bibitem{labelme}
``Image polygonal annotation with python,''
\url{https://github.com/wkentaro/labelme}.
\bibitem{yolov5}
ultralytics, ``yolov5,'' \url{https://github.com/ultralytics/yolov5}, 2022.
\bibitem{cao2021visdrone}
Y.~Cao, Z.~He, L.~Wang, W.~Wang, Y.~Yuan, D.~Zhang, J.~Zhang, P.~Zhu,
L.~Van~Gool, J.~Han \emph{et~al.}, ``Visdrone-det2021: The vision meets drone
object detection challenge results,'' in \emph{Proceedings of the IEEE/CVF
International Conference on Computer Vision}, 2021, pp. 2847--2854.
\bibitem{du2018unmanned}
D.~Du, Y.~Qi, H.~Yu, Y.~Yang, K.~Duan, G.~Li, W.~Zhang, Q.~Huang, and Q.~Tian,
``The unmanned aerial vehicle benchmark: Object detection and tracking,'' in
\emph{Proceedings of the European conference on computer vision (ECCV)},
2018, pp. 370--386.
\bibitem{cordts2016cityscapes}
M.~Cordts, M.~Omran, S.~Ramos, T.~Rehfeld, M.~Enzweiler, R.~Benenson,
U.~Franke, S.~Roth, and B.~Schiele, ``The cityscapes dataset for semantic
urban scene understanding,'' in \emph{Proceedings of the IEEE conference on
computer vision and pattern recognition}, 2016, pp. 3213--3223.
\bibitem{chen2019gated}
D.~Chen, M.~He, Q.~Fan, J.~Liao, L.~Zhang, D.~Hou, L.~Yuan, and G.~Hua, ``Gated
context aggregation network for image dehazing and deraining,'' in \emph{2019
IEEE winter conference on applications of computer vision (WACV)}.\hskip 1em
plus 0.5em minus 0.4em\relax IEEE, 2019, pp. 1375--1383.
\bibitem{qin2020ffa}
X.~Qin, Z.~Wang, Y.~Bai, X.~Xie, and H.~Jia, ``Ffa-net: Feature fusion
attention network for single image dehazing,'' in \emph{Proceedings of the
AAAI Conference on Artificial Intelligence}, vol.~34, no.~07, 2020, pp.
11\,908--11\,915.
\bibitem{chen2021srktdn}
T.~Chen, J.~Fu, W.~Jiang, C.~Gao, and S.~Liu, ``Srktdn: Applying super
resolution method to dehazing task,'' in \emph{Proceedings of the IEEE/CVF
Conference on Computer Vision and Pattern Recognition}, 2021, pp. 487--496.
\bibitem{fu2021dw}
M.~Fu, H.~Liu, Y.~Yu, J.~Chen, and K.~Wang, ``Dw-gan: A discrete wavelet
transform gan for nonhomogeneous dehazing,'' in \emph{Proceedings of the
IEEE/CVF Conference on Computer Vision and Pattern Recognition}, 2021, pp.
203--212.
\bibitem{liu2020trident}
J.~Liu, H.~Wu, Y.~Xie, Y.~Qu, and L.~Ma, ``Trident dehazing network,'' in
\emph{Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern
Recognition Workshops}, 2020, pp. 430--431.
\bibitem{engin2018cycle}
D.~Engin, A.~Gen{\c{c}}, and H.~Kemal~Ekenel, ``Cycle-dehaze: Enhanced cyclegan
for single image dehazing,'' in \emph{Proceedings of the IEEE conference on
computer vision and pattern recognition workshops}, 2018, pp. 825--833.
\bibitem{you2019ct}
C.~You, G.~Li, Y.~Zhang, X.~Zhang, H.~Shan, M.~Li, S.~Ju, Z.~Zhao, Z.~Zhang,
W.~Cong \emph{et~al.}, ``Ct super-resolution gan constrained by the
identical, residual, and cycle learning ensemble (gan-circle),'' \emph{IEEE
transactions on medical imaging}, vol.~39, no.~1, pp. 188--203, 2019.
\bibitem{zhu2017unpaired}
J.-Y. Zhu, T.~Park, P.~Isola, and A.~A. Efros, ``Unpaired image-to-image
translation using cycle-consistent adversarial networks,'' in
\emph{Proceedings of the IEEE international conference on computer vision},
2017, pp. 2223--2232.
\bibitem{johnson2016perceptual}
J.~Johnson, A.~Alahi, and L.~Fei-Fei, ``Perceptual losses for real-time style
transfer and super-resolution,'' in \emph{European conference on computer
vision}.\hskip 1em plus 0.5em minus 0.4em\relax Springer, 2016, pp. 694--711.
\bibitem{liu2020connecting}
D.~Liu, B.~Wen, J.~Jiao, X.~Liu, Z.~Wang, and T.~S. Huang, ``Connecting image
denoising and high-level vision tasks via deep learning,'' \emph{IEEE
Transactions on Image Processing}, vol.~29, pp. 3695--3706, 2020.
\end{thebibliography}