Filters








1,355 Hits in 3.6 sec

Polka Lines: Learning Structured Illumination and Reconstruction for Active Stereo [article]

Seung-Hwan Baek, Felix Heide
<span title="2021-05-26">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Such hand-crafted patterns are designed in isolation from the scene statistics, ambient illumination conditions, and the reconstruction method.  ...  In this work, we propose the first method to jointly learn structured illumination and reconstruction, parameterized by a diffractive optical element and a neural network, in an end-to-end fashion.  ...  Acknowledgements The authors are grateful to Ethan Tseng and Derek Nowrouzezahrai for fruitful discussions. Felix Heide was supported by an NSF CAREER Award (2047359) and a Sony Young Faculty Award.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2011.13117v2">arXiv:2011.13117v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/2biglpngxbayjgtkwho5j7fssi">fatcat:2biglpngxbayjgtkwho5j7fssi</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210607094141/https://arxiv.org/pdf/2011.13117v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/81/a4/81a4381176c4e01355bf03ca8147cd83ddae01d4.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2011.13117v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Learning to Think Outside the Box: Wide-Baseline Light Field Depth Estimation with EPI-Shift

Titus Leistner, Hendrik Schilling, Radek Mackowiak, Stefan Gumhold, Carsten Rother
<span title="">2019</span> <i title="IEEE"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/fnmyhnfycff7tiixnks6dqb6sy" style="color: black;">2019 International Conference on 3D Vision (3DV)</a> </i> &nbsp;
Our goal is to design a pipeline which achieves highly accurate results for small- and wide-baseline light fields.  ...  EPI-Shift considerably outperforms the state-of-the-art learning-based approaches and is on par with hand-crafted methods.  ...  We thank the Center for Information Services and High Performance Computing (ZIH) at TU Dresden for generous allocations of computer time.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/3dv.2019.00036">doi:10.1109/3dv.2019.00036</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/3dim/LeistnerSMGR19.html">dblp:conf/3dim/LeistnerSMGR19</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/eewtvohxjzexvhxd6zvhqugojm">fatcat:eewtvohxjzexvhxd6zvhqugojm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200823192737/https://arxiv.org/pdf/1909.09059v1.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/2f/30/2f3097e5a385533edaac43a731a0775148c922e1.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/3dv.2019.00036"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

Scale-Adaptive Neural Dense Features: Learning via Hierarchical Context Aggregation

Jaime Spencer, Richard Bowden, Simon Hadfield
<span title="">2019</span> <i title="IEEE"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/ilwxppn4d5hizekyd3ndvy2mii" style="color: black;">2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)</a> </i> &nbsp;
How do computers and intelligent agents view the world around them? Feature extraction and representation constitutes one the basic building blocks towards answering this question.  ...  These approaches, while successful, tend to lack scalability and can't easily exploit information learned by other systems.  ...  Acknowledgements This work was funded by the EPSRC under grant agreement (EP/R512217/1). We would also like to thank NVIDIA Corporation for their Titan Xp GPU grant.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/cvpr.2019.00636">doi:10.1109/cvpr.2019.00636</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/cvpr/SpencerBH19.html">dblp:conf/cvpr/SpencerBH19</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/6z2csmpfkbfuljcszrdnqexwne">fatcat:6z2csmpfkbfuljcszrdnqexwne</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200505082645/http://epubs.surrey.ac.uk/850894/1/Scale-Adaptive%20Neural%20Dense%20Features.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/09/1b/091b0ad82f3900ddd301b3a2098bccc5a9d75f0c.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/cvpr.2019.00636"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

Scale-Adaptive Neural Dense Features: Learning via Hierarchical Context Aggregation [article]

Jaime Spencer, Richard Bowden, Simon Hadfield
<span title="2019-03-25">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
How do computers and intelligent agents view the world around them? Feature extraction and representation constitutes one the basic building blocks towards answering this question.  ...  These approaches, while successful, tend to lack scalability and can't easily exploit information learned by other systems.  ...  Acknowledgements This work was funded by the EPSRC under grant agreement (EP/R512217/1). We would also like to thank NVIDIA Corporation for their Titan Xp GPU grant.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1903.10427v1">arXiv:1903.10427v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/hqnnzdfczjhbjmpbj6surelo6i">fatcat:hqnnzdfczjhbjmpbj6surelo6i</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20191021060743/https://arxiv.org/pdf/1903.10427v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/f5/be/f5be36a2deea516c82206e89f6158ecfea80342e.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1903.10427v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

UltraStereo: Efficient Learning-Based Matching for Active Stereo Systems

Sean Ryan Fanello, Julien Valentin, Christoph Rhemann, Adarsh Kowdle, Vladimir Tankovich, Philip Davidson, Shahram Izadi
<span title="">2017</span> <i title="IEEE"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/ilwxppn4d5hizekyd3ndvy2mii" style="color: black;">2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)</a> </i> &nbsp;
We efficiently solve the specialized problem of stereo matching under active illumination using a new learning-based algorithm.  ...  Extensive evaluations show we surpass the quality and overcome the limitations of current depth sensing technologies.  ...  Acknowledgements We thank the entire perceptiveIO team for continuous feedback and support regarding this work.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/cvpr.2017.692">doi:10.1109/cvpr.2017.692</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/cvpr/FanelloVRKTDI17.html">dblp:conf/cvpr/FanelloVRKTDI17</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/auf6kvbp5nfabi2glqbcovpiva">fatcat:auf6kvbp5nfabi2glqbcovpiva</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20171209085248/http://openaccess.thecvf.com:80/content_cvpr_2017/papers/Fanello_UltraStereo_Efficient_Learning-Based_CVPR_2017_paper.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/a8/b7/a8b7a6aff04016b394426c91f6b98e47961a6ee0.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/cvpr.2017.692"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

Stereopagnosia: Fooling Stereo Networks with Adversarial Perturbations [article]

Alex Wong, Mukund Mundhra, Stefano Soatto
<span title="2021-03-26">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We study the effect of adversarial perturbations of images on the estimates of disparity by deep learning models trained for stereo.  ...  We test our method using the most recent stereo networks and evaluate their performance on public benchmark datasets.  ...  Ethical Impact As deep learning models are widely deployed for various tasks, adversarial examples have been treated as a threat to the security of such applications.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2009.10142v3">arXiv:2009.10142v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/q4dpuxw55ndhhnmsc3hekqda6a">fatcat:q4dpuxw55ndhhnmsc3hekqda6a</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200924001600/https://arxiv.org/pdf/2009.10142v1.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2009.10142v3" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

TW-SMNet: Deep Multitask Learning of Tele-Wide Stereo Matching [article]

Mostafa El-Khamy, Haoyu Ren, Xianzhi Du, Jungwon Lee
<span title="2019-06-11">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Moreover, we design multiple methods for the fusion of the SMDE and SIDE networks.  ...  We further propose an end-to-end deep multitask tele-wide stereo matching neural network (MT-TW-SMNet), which simultaneously learns the SMDE task for the overlapped Tele FOV and the single image inverse  ...  We consider multiple deep-learning based approaches to the tele-wide stereo matching problem.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1906.04463v1">arXiv:1906.04463v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/sdqb3pjqc5ag3cxb2f46erhaaq">fatcat:sdqb3pjqc5ag3cxb2f46erhaaq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20191222094931/https://arxiv.org/pdf/1906.04463v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/3a/55/3a55ad9c3e2c80774aaf5e7168bb08d96beb18c7.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1906.04463v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Discriminative Learning of Deep Convolutional Feature Point Descriptors

Edgar Simo-Serra, Eduard Trulls, Luis Ferraz, Iasonas Kokkinos, Pascal Fua, Francesc Moreno-Noguer
<span title="">2015</span> <i title="IEEE"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/753trptklbb4nj6jquqadzwwdu" style="color: black;">2015 IEEE International Conference on Computer Vision (ICCV)</a> </i> &nbsp;
Deep learning has revolutionalized image-level tasks such as classification, but patch-level tasks, such as correspondence, still rely on hand-crafted features, e.g. SIFT.  ...  By using the L 2 distance during both training and testing we develop 128-D descriptors whose euclidean distances reflect patch similarity, and which can be used as a drop-in replacement for any task involving  ...  Our models generalize well across different datasets and applications, including wide-baseline matching, non-rigid deformations and extreme illumination changes.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/iccv.2015.22">doi:10.1109/iccv.2015.22</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/iccv/Simo-SerraTFKFM15.html">dblp:conf/iccv/Simo-SerraTFKFM15</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/s7ovndlftjf4rapj2cxt65llwe">fatcat:s7ovndlftjf4rapj2cxt65llwe</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20170818222642/https://infoscience.epfl.ch/record/213228/files/iccv-2015-deepdesc.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/84/39/84394748e2af0121ee1038800679f9cecc8dff23.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/iccv.2015.22"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

Review of Wide-Baseline Stereo Image Matching Based on Deep Learning

Guobiao Yao, Alper Yilmaz, Fei Meng, Li Zhang
<span title="2021-08-17">2021</span> <i title="MDPI AG"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/kay2tsbijbawliu45dnhvyvgsq" style="color: black;">Remote Sensing</a> </i> &nbsp;
We present the results of comprehensive experiments on actual wide-baseline stereo images, which we use to contrast and discuss the advantages and disadvantages of several state-of-the-art deep-learning  ...  Deep convolutional neural networks, which can express high-level features and their correlation, have received increasing attention for the task of wide-baseline image matching, and learning-based methods  ...  We are very grateful also for the valuable comments and contributions of anonymous reviewers and members of the editorial team. Conflicts of Interest: The authors declare no conflict of interest.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3390/rs13163247">doi:10.3390/rs13163247</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/u43b2al6dzblddofc4ggcamiki">fatcat:u43b2al6dzblddofc4ggcamiki</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211210205643/https://mdpi-res.com/d_attachment/remotesensing/remotesensing-13-03247/article_deploy/remotesensing-13-03247.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/56/c0/56c0184d405004e099691c53c81afcd9b6520d25.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3390/rs13163247"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> mdpi.com </button> </a>

Multitask Deep Neural Networks for Tele-Wide Stereo Matching

Mostafa El-Khamy, Haoyu Ren, Xianzhi Du, Jungwon Lee
<span title="">2020</span> <i title="Institute of Electrical and Electronics Engineers (IEEE)"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/q7qi7j4ckfac7ehf3mjbso4hne" style="color: black;">IEEE Access</a> </i> &nbsp;
In this paper, we consider multiple deep-learning solutions to the tele-wide stereo matching problem.  ...  Hence, multiple disparity measurements for the common (overlapping) field of view are estimated by stereo pairs with multiple baselines, and then fused to obtain a statistically more accurate depth map  ...  He has co-authored more than 130 papers and holds over 300 patents in the US and many more worldwide. Dr. Lee is an IEEE Fellow.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/access.2020.3029085">doi:10.1109/access.2020.3029085</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/jlqgl6cbbveblnx5jdjztehqpa">fatcat:jlqgl6cbbveblnx5jdjztehqpa</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201007065224/https://ieeexplore.ieee.org/ielx7/6287639/6514899/09214409.pdf?tp=&amp;arnumber=9214409&amp;isnumber=6514899&amp;ref=" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/1f/ef/1fefef44b4ccefc121b4b907ebe9d07af551a168.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/access.2020.3029085"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> ieee.com </button> </a>

Learning to compare image patches via convolutional neural networks

Sergey Zagoruyko, Nikos Komodakis
<span title="">2015</span> <i title="IEEE"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/ilwxppn4d5hizekyd3ndvy2mii" style="color: black;">2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)</a> </i> &nbsp;
To that end, we explore and study multiple neural network architectures, which are specifically adapted to this task.  ...  We show that such an approach can significantly outperform the state-ofthe-art on several problems and benchmark datasets.  ...  For the rest of the experiments, we note that we use models trained on the Liberty dataset. Wide baseline stereo evaluation For this evaluation we chose the dataset by Strecha et al.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/cvpr.2015.7299064">doi:10.1109/cvpr.2015.7299064</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/cvpr/ZagoruykoK15.html">dblp:conf/cvpr/ZagoruykoK15</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/vpwke4ad6bhixg2puzbdbw4zhq">fatcat:vpwke4ad6bhixg2puzbdbw4zhq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20150906065328/http://www.cv-foundation.org:80/openaccess/content_cvpr_2015/papers/Zagoruyko_Learning_to_Compare_2015_CVPR_paper.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/d6/dc/d6dcb3c2e69db84de75f1f5b6e47ba33bca2a602.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/cvpr.2015.7299064"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

EdgeStereo: An Effective Multi-Task Learning Network for Stereo Matching and Edge Detection [article]

Xiao Song, Xu Zhao, Liangji Fang, Hanwen Hu
<span title="2019-12-11">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
To effectively incorporate edge cues, we propose the edge-aware smoothness loss and edge feature embedding for inter-task interactions.  ...  By the time of the paper submission, EdgeStereo achieves state-of-art performance on the FlyingThings3D dataset, KITTI 2012 and KITTI 2015 stereo benchmarks, outperforming other published stereo matching  ...  Acknowledgements We would like to thank Guorun Yang for helpful discussions and suggestions. This research is supported by the funding from NSFC programs (61673269, 61273285, U1764264).  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1903.01700v2">arXiv:1903.01700v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/st4jqj5gmzdjljh623d2xpmiwq">fatcat:st4jqj5gmzdjljh623d2xpmiwq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200902033639/https://arxiv.org/pdf/1903.01700v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/22/2a/222ae95568efe398e558ff5dd40ac8911476f8a2.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1903.01700v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

A Generalized Multi-Task Learning Approach to Stereo DSM Filtering in Urban Areas [article]

Lukas Liebel, Ksenia Bittner, Marco Körner
<span title="2020-04-07">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
However, stereo DSMs often suffer from noise and blur. Furthermore, they are heavily distorted by vegetation, which is of lesser relevance for most applications.  ...  Our encoder-decoder models with shared encoders and multiple task-specific decoders leverage roof type classification as a secondary task and multiple objectives including a conditional adversarial term  ...  Acknowledgements The work of Lukas Liebel was funded by the German Federal Ministry of Transport and Digital Infrastructure (BMVI) under reference 16AVF2019A.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2004.02493v2">arXiv:2004.02493v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/7losmhouibcxhk5qf7iobuvmcq">fatcat:7losmhouibcxhk5qf7iobuvmcq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200410223010/https://arxiv.org/pdf/2004.02493v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2004.02493v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

A Light Dual-Task Neural Network for Haze Removal

Yu Zhang, Xinchao Wang, Xiaojun Bi, Dacheng Tao
<span title="">2018</span> <i title="Institute of Electrical and Electronics Engineers (IEEE)"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/msfmoh6v7bdk7lrsmtbklto74i" style="color: black;">IEEE Signal Processing Letters</a> </i> &nbsp;
In LDTNet, the haze-free image and the transmission map are produced simultaneously. As a result, the artificial prior is reduced to the smallest extent.  ...  We use transmission map estimation as an auxiliary task to assist the main task, haze removal, in feature extraction and to enhance the generalization of the network.  ...  mean PSNR and SSIM on the Middlebury stereo dataset.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/lsp.2018.2849681">doi:10.1109/lsp.2018.2849681</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/3zirwmvblbcfdblcq5mfbflaku">fatcat:3zirwmvblbcfdblcq5mfbflaku</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200924152921/https://arxiv.org/pdf/1904.06024v1.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/ee/c8/eec87374a4146eb4b7cbfee6c90f91da0bef7371.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/lsp.2018.2849681"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

Working hard to know your neighbor's margins: Local descriptor learning loss [article]

Anastasiya Mishchuk, Dmytro Mishkin, Filip Radenovic, Jiri Matas
<span title="2018-01-12">2018</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Applying the novel loss to the L2Net CNN architecture results in a compact descriptor -- it has the same dimensionality as SIFT (128) that shows state-of-art performance in wide baseline stereo, patch  ...  We introduce a novel loss for learning local feature descriptors which is inspired by the Lowe's matching criterion for SIFT.  ...  Acknowledgements The authors were supported by the Czech Science Foundation Project GACR P103/12/G084, the Austrian Ministry for Transport, Innovation and Technology, the Federal Ministry of Science, Research  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1705.10872v4">arXiv:1705.10872v4</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/hdypqwvdmjccdhjhfktqrdsjyu">fatcat:hdypqwvdmjccdhjhfktqrdsjyu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200825010742/https://arxiv.org/pdf/1705.10872v4.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/00/d2/00d2e3b8d882aa0735aaa914fb6c4c7bddf88043.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1705.10872v4" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>
&laquo; Previous Showing results 1 &mdash; 15 out of 1,355 results