Filters








38,796 Hits in 2.3 sec

Anchor Box Optimization for Object Detection [article]

Yuanyi Zhong, Jianfeng Wang, Jian Peng, Lei Zhang
<span title="2020-01-26">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In this paper, we propose a general approach to optimize anchor boxes for object detection. Nowadays, anchor boxes are widely adopted in state-of-the-art detection frameworks.  ...  The learning approach can be easily implemented with stochastic gradient descent and can be plugged into any anchor box-based detection framework.  ...  Implementation Details Since the proposed approach for optimizing anchors is quite general, it can be applied to most anchor-based object detection frameworks.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1812.00469v2">arXiv:1812.00469v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/b3sfka5jwbf5lemlagsnxlnej4">fatcat:b3sfka5jwbf5lemlagsnxlnej4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200321080040/https://arxiv.org/pdf/1812.00469v2.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/1812.00469v2" 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>

AABO: Adaptive Anchor Box Optimization for Object Detection via Bayesian Sub-sampling [article]

Wenshuo Ma, Tingzhong Tian, Hang Xu, Yimin Huang, Zhenguo Li
<span title="2020-07-18">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
However, this might be sub-optimal or even wrong when a new dataset or a new model is adopted. In this paper, we study the problem of automatically optimizing anchor boxes for object detection.  ...  We first demonstrate that the number of anchors, anchor scales and ratios are crucial factors for a reliable object detection system.  ...  anchor boxes can improve the performance of object detection systems.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2007.09336v1">arXiv:2007.09336v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/ome26u2g7nfvjl7wimgckwysxi">fatcat:ome26u2g7nfvjl7wimgckwysxi</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200808173412/https://arxiv.org/pdf/2007.09336v1.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/2007.09336v1" 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>

Multiple Anchor Learning for Visual Object Detection [article]

Wei Ke and Tianliang Zhang and Zeyi Huang and Qixiang Ye and Jianzhuang Liu and Dong Huang
<span title="2019-12-04">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
However, in CNN-based detectors, these two modules are usually optimized under a fixed set of candidate (or anchor) bounding boxes.  ...  In this paper, we propose a Multiple Instance Learning (MIL) approach that selects anchors and jointly optimizes the two modules of a CNN-based object detector.  ...  FCOS leverages cell-level supervision and center-ness bounding-box regression [29] for object detection.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1912.02252v1">arXiv:1912.02252v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/nwe35ue2nfhmjpbg772weinvtq">fatcat:nwe35ue2nfhmjpbg772weinvtq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200922214052/https://arxiv.org/pdf/1912.02252v1.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/f7/e1/f7e17660e34b784a49955aa6e036481856765853.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1912.02252v1" 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 Robust Thermal Infrared Vehicle and Pedestrian Detection Method in Complex Scenes

Yang Liu, Hailong Su, Cao Zeng, Xiaoli Li
<span title="2021-02-10">2021</span> <i title="MDPI AG"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/taedaf6aozg7vitz5dpgkojane" style="color: black;">Sensors</a> </i> &nbsp;
inference, which efficiently improves the detection precision of the detector for tiny and dense objects.  ...  to enhance the detection performance of motion-blurred objects.  ...  However, this method has poor detection performance for tiny objects that are difficult to model based on anchor boxes, and the object recall is low. Zhang et al.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3390/s21041240">doi:10.3390/s21041240</a> <a target="_blank" rel="external noopener" href="https://www.ncbi.nlm.nih.gov/pubmed/33578700">pmid:33578700</a> <a target="_blank" rel="external noopener" href="https://pubmed.ncbi.nlm.nih.gov/PMC7916389/">pmcid:PMC7916389</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/7amjv5wqrvdypozpmz7mljx4o4">fatcat:7amjv5wqrvdypozpmz7mljx4o4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210214052922/https://res.mdpi.com/d_attachment/sensors/sensors-21-01240/article_deploy/sensors-21-01240.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/7d/64/7d64e94f8b8ddb494db95d12855f8cbf95036728.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3390/s21041240"> <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> <a target="_blank" rel="external noopener" href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7916389" title="pubmed link"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> pubmed.gov </button> </a>

Black-Box Optimization of Object Detector Scales [article]

Mohandass Muthuraja and Octavio Arriaga and Paul Plöger and Frank Kirchner and Matias Valdenegro-Toro
<span title="2020-10-29">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In this work, we propose the use of Black-box optimization methods to tune the prior/default box scales in Faster R-CNN and SSD, using Bayesian Optimization, SMAC, and CMA-ES.  ...  We show that by tuning the input image size and prior box anchor scale on Faster R-CNN mAP increases by 2% on PASCAL VOC 2007, and by 3% with SSD.  ...  Object Detection Hyper-parameters Selection The designing of prior/default anchor boxes is a big challenge in object detection.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2010.15823v1">arXiv:2010.15823v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/ab3yxoigxrcu3isrg42hhmr33i">fatcat:ab3yxoigxrcu3isrg42hhmr33i</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201031153133/https://arxiv.org/pdf/2010.15823v1.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/e3/df/e3df72800e867a63dc4e2c8da85fca2f798b88bd.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2010.15823v1" 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>

Dynamic Anchor: A Feature-Guided Anchor Strategy for Object Detection

Xing Liu, Huai-Xin Chen, Bi-Yuan Liu
<span title="2022-05-12">2022</span> <i title="MDPI AG"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/smrngspzhzce7dy6ofycrfxbim" style="color: black;">Applied Sciences</a> </i> &nbsp;
The majority of modern object detectors rely on a set of pre-defined anchor boxes, which enhances detection performance dramatically.  ...  In addition, only one anchor box is predicted for each location, which dramatically reduces calculation.  ...  As the optimal anchor box of an object is related to its features, the detectors attempt to predict the shapes of anchor boxes with feature maps.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3390/app12104897">doi:10.3390/app12104897</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/isqlkojiqncgtdlmpy5lybtreu">fatcat:isqlkojiqncgtdlmpy5lybtreu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220514210301/https://mdpi-res.com/d_attachment/applsci/applsci-12-04897/article_deploy/applsci-12-04897.pdf?version=1652347227" 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/8d/0e/8d0e2f9c470a0f40596728aa0f51196cd4cb0776.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3390/app12104897"> <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>

Anchor Pruning for Object Detection [article]

Maxim Bonnaerens, Matthias Freiberger, Joni Dambre
<span title="2021-07-20">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
This paper proposes anchor pruning for object detection in one-stage anchor-based detectors.  ...  In this work we demonstrate an additional pruning technique, specifically for object detection: anchor pruning.  ...  CornerNet (Law and Deng (2018) ) for example, detects corners of an object and groups multiple predictions together to produce a final bounding box.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2104.00432v2">arXiv:2104.00432v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/hm4ts6jvrvfd7b3sz2otygcyhi">fatcat:hm4ts6jvrvfd7b3sz2otygcyhi</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210722115126/https://arxiv.org/pdf/2104.00432v2.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/22/81/2281dbf27bc858e950c26b8147af04599073d7c9.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2104.00432v2" 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>

Consistent Optimization for Single-Shot Object Detection [article]

Tao Kong, Fuchun Sun, Huaping Liu, Yuning Jiang, Jianbo Shi
<span title="2019-01-23">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We present consistent optimization for single stage object detection.  ...  Previous works of single stage object detectors usually rely on the regular, dense sampled anchors to generate hypothesis for the optimization of the model.  ...  Detection model that is optimized only based on the original anchors is loose and inaccurate.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1901.06563v2">arXiv:1901.06563v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/ztznizqzczhrholve5bpyn4q4m">fatcat:ztznizqzczhrholve5bpyn4q4m</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200829122206/https://arxiv.org/pdf/1901.06563v2.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/ba/5f/ba5fcc1a691ecb0e7bc917eb1c00b5fd01701fab.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1901.06563v2" 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>

Localize to Classify and Classify to Localize: Mutual Guidance in Object Detection [article]

Heng Zhang, Elisa Fromont, Sébastien Lefevre, Bruno Avignon
<span title="2020-09-29">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Most deep learning object detectors are based on the anchor mechanism and resort to the Intersection over Union (IoU) between predefined anchor boxes and ground truth boxes to evaluate the matching quality  ...  between anchors and objects.  ...  Anchor-free object detection The idea of anchor-free object detection consists in detecting objects not from predefined anchors boxes, but directly from particular key-points [18, 19, 20, 21] or object  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2009.14085v1">arXiv:2009.14085v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/lc6mo6wgbzdtdbnqe3lekznj7a">fatcat:lc6mo6wgbzdtdbnqe3lekznj7a</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201001050954/https://arxiv.org/pdf/2009.14085v1.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/2009.14085v1" 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>

FoveaBox: Beyond Anchor-based Object Detector [article]

Tao Kong, Fuchun Sun, Huaping Liu, Yuning Jiang, Jianbo Shi
<span title="2019-04-08">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We present FoveaBox, an accurate, flexible and completely anchor-free framework for object detection.  ...  Instead, FoveaBox directly learns the object existing possibility and the bounding box coordinates without anchor reference.  ...  By simultaneously predict the object position and the corresponding boundary, FoveaBox gives a clean solution for detecting objects without prior candidate boxes.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1904.03797v1">arXiv:1904.03797v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/tp5r2nyzone4bjbnbnr4z6p64i">fatcat:tp5r2nyzone4bjbnbnr4z6p64i</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200929205113/https://arxiv.org/pdf/1904.03797v1.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/e3/63/e363ee94fb5b1061a813a47faf40b833eb5de233.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1904.03797v1" 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>

Circle Representation for Medical Object Detection [article]

Ethan H. Nguyen, Haichun Yang, Ruining Deng, Yuzhe Lu, Zheyu Zhu, Joseph T. Roland, Le Lu, Bennett A. Landman, Agnes B. Fogo, Yuankai Huo
<span title="2021-10-22">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Box representation has been extensively used for object detection in computer vision.  ...  In this paper, we propose a simple circle representation for medical object detection and introduce CircleNet, an anchor-free detection framework.  ...  Nguyen thanks Jonathan Ehrman for hardware support and Rose Rasty for editing and proofreading.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2110.12093v1">arXiv:2110.12093v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/ommwz5gcqvgoxhz2myypy5mpii">fatcat:ommwz5gcqvgoxhz2myypy5mpii</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211102142041/https://arxiv.org/pdf/2110.12093v1.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/e3/29/e329be3641ac360d7c3f4ce67033443db4f4cfdb.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2110.12093v1" 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 New Object Detection Method for Object Deviating from Center or Multi Object Crowding

Yubo Liang, Gongming Wang, Wenfa Li, Yuzhe He, Xiaoming Liang
<span title="">2021</span> <i title="Elsevier BV"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/ne2g4brfavewtlp6ncjux7lwxa" style="color: black;">Displays (Guildford)</a> </i> &nbsp;
To solve this problem, we use a variety of optimization methods for RetinaNet to improve the accuracy of object detection.  ...  Highlights 1) Proposing an object detection method for object deviating from center or multi object crowding. 2) The functions of PA-FPN, FreeAnchor and Soft-NMS are complementary and the combination of  ...  For using the object detection algorithm based on anchor,a set of fixed anchor box with different sizes and positions should be preset, which can cover all sizes and positions.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1016/j.displa.2021.102042">doi:10.1016/j.displa.2021.102042</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/d7gvbsrqv5fvbcntx74lgnre4q">fatcat:d7gvbsrqv5fvbcntx74lgnre4q</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210701060809/https://pdf.sciencedirectassets.com/271537/AIP/1-s2.0-S0141938221000536/main.pdf?X-Amz-Security-Token=IQoJb3JpZ2luX2VjEN7%2F%2F%2F%2F%2F%2F%2F%2F%2F%2FwEaCXVzLWVhc3QtMSJHMEUCIFepzkCPJvBOJ7V2IXGK3KbqiDTkA%2BdDiHVJSJsXLOgsAiEA9OBxmiddT9kXIMzV0P8L8ivHMHb7axO5CU4S6FeOAfYqgwQItv%2F%2F%2F%2F%2F%2F%2F%2F%2F%2FARAEGgwwNTkwMDM1NDY4NjUiDPAz%2B1GvIRjH7U%2FZqCrXA0ND%2BCuJGeRoWecV4yKdimpPYN7UmTtQP5pmonCb6bjN4prB1yBye9CsrbNhNNKuPLiOzqFOfH7cuUP1gV5B5EqBmZO0l1wej5ej0GKraRbjh0gPUAALBdN4zUJaKi%2B4OaD1lc3S%2FwCmRGxWWEj9jqEFTXkaE8M89x5g830hEaBG9gxjeSmfFi6zHVf8zdF%2FQz1Vv5hYgwon5U14h3kipST506Hou3ecjpwi0DxnjY5Erml%2F6Q5RQqOdu5JfPrpQN4a41XUg9UYxgE45i7lbMVEUqqYrTnqmFMoIsF5vS2zXnHWGkDYXkL%2BuMNfMf0PwqfgSyUQ9579doG27bntQM4B92FLW3UICEonll5SOa6St%2FFVkznHWUIfdxjQKIL0pyJ05oSuRlaemHz%2BWW6PDQ9IbNZfO4Km4t%2BKWh3HlEv5xs98dTlRtAtSvJfWCiDA6BY3i9%2FaDDxPhvIybQkgTx%2F27T7nB%2FcP7JM5CCPQXR9qLwuVl5xVuKB069XrxmSuBiDyk13zAGc4wxXPagKKLk%2BJ%2BR7k4BRN6hcXGGbUlXDTtUtY5VpBE8uQw8G5eStZbJl59KIe2zkb7qW4H5UQybrxG9Fel7X4OHtOMLJfzRCH2plz2qEWETTCoofWGBjqlAYzLMtVtqDEMv5Gq91kxdboQEcUCNsezOF4VExhUu7nSyK4b3jXzgnRFKvhgVZYHbOijDiFR3KvUOwiDdTetX3tLUw4HM%2BI0RHY13V43BVoxK08O1XP8uuXVVpzj7MQGs0j9hjNxXikIX5fspOb6xmnNhIwKAMn4ftnTRFa0vLKh%2B0DIqM3g0%2BFxmWo2%2ByOmAUFxM5XXrz%2F8QFhJLC1CXrafrufucQ%3D%3D&amp;X-Amz-Algorithm=AWS4-HMAC-SHA256&amp;X-Amz-Date=20210701T060803Z&amp;X-Amz-SignedHeaders=host&amp;X-Amz-Expires=300&amp;X-Amz-Credential=ASIAQ3PHCVTYXYD6JBEQ%2F20210701%2Fus-east-1%2Fs3%2Faws4_request&amp;X-Amz-Signature=f966f4965a97d298fce9137d1d394fbc4244aa67d11906f67bb4d873298debf8&amp;hash=ac88b0e5213db1e5921c00d25162763bc805fb8ab4311619ed92b9a8e3e3e545&amp;host=68042c943591013ac2b2430a89b270f6af2c76d8dfd086a07176afe7c76c2c61&amp;pii=S0141938221000536&amp;tid=spdf-153fa218-7e00-4ee6-825b-1d47a9ca532d&amp;sid=6c725cc09c3da54be669d203918c81004604gxrqa&amp;type=client" 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/df/40/df407a45429d6700bb768748b0a58b26e06e237e.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1016/j.displa.2021.102042"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> elsevier.com </button> </a>

UAV-YOLO: Small Object Detection on Unmanned Aerial Vehicle Perspective

Mingjie Liu, Xianhao Wang, Anjian Zhou, Xiuyuan Fu, Yiwei Ma, Changhao Piao
<span title="2020-04-15">2020</span> <i title="MDPI AG"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/taedaf6aozg7vitz5dpgkojane" style="color: black;">Sensors</a> </i> &nbsp;
In this study, the authors develop a special detection method for small objects in UAV perspective.  ...  Both these two optimizations can enlarge the receptive filed. Furthermore, UAV-viewed dataset is collected to UAV perspective or small object detection.  ...  Optimizing the number and size of anchor box by k-means. As mentioned before, we also optimized the number and size of anchor box by k-means.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3390/s20082238">doi:10.3390/s20082238</a> <a target="_blank" rel="external noopener" href="https://www.ncbi.nlm.nih.gov/pubmed/32326573">pmid:32326573</a> <a target="_blank" rel="external noopener" href="https://pubmed.ncbi.nlm.nih.gov/PMC7218847/">pmcid:PMC7218847</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/ukc4erzm3zg3reziqcxqq56t4y">fatcat:ukc4erzm3zg3reziqcxqq56t4y</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200505164913/https://res.mdpi.com/d_attachment/sensors/sensors-20-02238/article_deploy/sensors-20-02238-v2.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/86/a2/86a241a0baaac35851f280e5d8de13728149bd82.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3390/s20082238"> <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> <a target="_blank" rel="external noopener" href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7218847" title="pubmed link"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> pubmed.gov </button> </a>

Deep Window Detection in Street Scenes

<span title="2020-02-29">2020</span> <i title="Korean Society for Internet Information (KSII)"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/hupfbobgkvepdnt5g32qxkypsy" style="color: black;">KSII Transactions on Internet and Information Systems</a> </i> &nbsp;
Recently, convolutional neural networks based object detection algorithms attract much attention due to their good performances.  ...  Unfortunately, directly training them for challenging window detection cannot achieve satisfying results. In this paper, we propose an approach for window detection.  ...  One-stage detectors, such as RetinaNet [17] , uses anchors to regress the bounding boxes of objects directly. For window detection, we design a special anchor configuration for the window detector.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3837/tiis.2020.02.022">doi:10.3837/tiis.2020.02.022</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/5gtaup5tbnhg5l464hqnj3prti">fatcat:5gtaup5tbnhg5l464hqnj3prti</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200505234645/http://itiis.org/digital-library/manuscript/file/23281/TIIS%20Vol%2014,%20No%202-22.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/c4/14/c414867bd7e6254e54eae6dbd0ff12eb50f2f5d7.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.3837/tiis.2020.02.022"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>

Feature Learning Improved by Location Guidance and Supervision for Object Detection

Bingying Li, Jiale Xiong, Xiang Fu, Jiexian Zeng, Lu Leng
<span title="">2021</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;
Therefore, two-stage detection can obtain more accurate object features by optimizing the anchor box.  ...  The detector first recognizes the approximate location of the object in the image, and then the optimized anchor box guides ROIPooling to extract object features.  ...  Leng is a member of the Institute of Electrical and Electronics Engineers (IEEE), Association for Computing Machinery (ACM), China Society of Image and Graphics (CSIG) and China Computer Federation (CCF  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/access.2021.3110888">doi:10.1109/access.2021.3110888</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/vc2uviow5rdrfgc57oekblay2y">fatcat:vc2uviow5rdrfgc57oekblay2y</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210917034452/https://ieeexplore.ieee.org/ielx7/6287639/6514899/09530548.pdf?tp=&amp;arnumber=9530548&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/1e/66/1e666bd14da8fe19f33e9e24f634288ecbeaccad.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/access.2021.3110888"> <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>
&laquo; Previous Showing results 1 &mdash; 15 out of 38,796 results