Filters








13,868 Hits in 6.6 sec

Deep Residual Transform for Multi-scale Image Decomposition

Yuhao Chen, Alexander Wong, Yuan Fang, Yifan Wu, Linlin Xu
<span title="2021-01-15">2021</span> <i title="University of Waterloo"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/fdnvq5cm6vdnxhlmdvxkkmcj6q" style="color: black;">Journal of Computational Vision and Imaging Systems</a> </i> &nbsp;
levels of visual granularity from coarse structures to fine details.  ...  Multi-scale image decomposition (MID) is a fundamental task in computer vision and image processing that involves the transformation of an image into a hierarchical representation comprising of different  ...  In particular, the learning progression of a 2-layer autoencoder (as shown in 4 top row) describes a non-linear transition from coarse structures to fine details for the representation, whereas the learning  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.15353/jcvis.v6i1.3537">doi:10.15353/jcvis.v6i1.3537</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/fr2ij5exqffohi7xxjnjxpdzl4">fatcat:fr2ij5exqffohi7xxjnjxpdzl4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210209113251/https://openjournals.uwaterloo.ca/index.php/vsl/article/download/3537/4583" 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/17/22/1722246e2b63cab8e62512fa99c410849fc24417.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.15353/jcvis.v6i1.3537"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>

Multi-Neighborhood Convolutional Networks

Elnaz Barshan, Paul Fieguth, Alexander Wong
<span title="2015-10-31">2015</span> <i title="University of Waterloo"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/ezsxysndlvgxhd4frt2jhaaeiq" style="color: black;">Vision Letters</a> </i> &nbsp;
We propose multi-neighborhood convolutional<br />networks, designed to learn image features at different levels of<br />detail.  ...  <p>We explore the role of scale for improved feature learning in convolutional<br />networks.  ...  we need a multi-neighborhood architecture with small/large neighborhoods at fine/coarse scales, respectively.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.15353/vsnl.v1i1.56">doi:10.15353/vsnl.v1i1.56</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/uoshjob7dbcqdkq4lhqemb4q44">fatcat:uoshjob7dbcqdkq4lhqemb4q44</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20170921235234/http://openjournals.uwaterloo.ca/index.php/vsl/article/download/VL122/48" 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/92/c9/92c950c397cbe246ddf04b0d0be90ecab27de6de.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.15353/vsnl.v1i1.56"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>

Multi-level 3D CNN for Learning Multi-scale Spatial Features [article]

Sambit Ghadai, Xian Lee, Aditya Balu, Soumik Sarkar, Adarsh Krishnamurthy
<span title="2019-05-03">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
The performance of our multi-level learning algorithm for object recognition is comparable to dense voxel representations while using significantly lower memory.  ...  To demonstrate the utility of the proposed multi-level learning, we use a multi-level voxel representation of 3D objects to perform object recognition.  ...  Hence, efficient and scalable deep learning techniques that exploit sparse and hierarchical data representations are necessary to deal with large 3D data sets.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1805.12254v2">arXiv:1805.12254v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/xfmichbn2jgn7gllu2uyknfkzu">fatcat:xfmichbn2jgn7gllu2uyknfkzu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20191013204940/https://arxiv.org/pdf/1805.12254v2.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/cd/f8/cdf8a980cfd75f3e6a13f1a6496b7c1d321b3107.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1805.12254v2" 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>

Hybrid coarse-fine classification for head pose estimation [article]

Haofan Wang, Zhenghua Chen, Yi Zhou
<span title="2019-10-02">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In this paper, to do the estimation without facial landmarks, we combine the coarse and fine regression output together for a deep network.  ...  Utilizing more quantization units for the angles, a fine classifier is trained with the help of other auxiliary coarse units. Integrating regression is adopted to get the final prediction.  ...  We take both coarse bin classification and relatively fine bin classification into account, each FC layer represents a different classification scale and compute its own cross-entropy loss.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1901.06778v2">arXiv:1901.06778v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/iydw3t2sg5dwtnc3wn4h6jpj2y">fatcat:iydw3t2sg5dwtnc3wn4h6jpj2y</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200826121253/https://arxiv.org/pdf/1901.06778v2.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/93/e2/93e2e15d5ca695c229f1287dec78b99bfab56003.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1901.06778v2" 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 scale-variant and scale-invariant features for deep image classification [article]

Nanne van Noord, Eric Postma
<span title="2016-05-13">2016</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Convolutional Neural Networks (CNNs) require large image corpora to be trained on classification tasks.  ...  We propose a multi- scale CNN method to encourage the recognition of both types of features and evaluate it on a challenging image classification task involving task-relevant characteristics at multiple  ...  However, using a multi-scale approach to artist attribution it is possible to use information from different scales, learning features appropriate from both coarse and fine details. V.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1602.01255v2">arXiv:1602.01255v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/4rou6ytpozaynfjcoy2b376vgy">fatcat:4rou6ytpozaynfjcoy2b376vgy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200826220702/https://arxiv.org/pdf/1602.01255v2.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/97/da/97da965a6f147767608a8b4435398d9f9dbe6089.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1602.01255v2" 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>

Stacked Semantic-Guided Network for Zero-Shot Sketch-Based Image Retrieval [article]

Hao Wang, Cheng Deng, Xinxu Xu, Wei Liu, Xinbo Gao, Dacheng Tao
<span title="2019-10-18">2019</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Specifically, we devise multi-layer feature fusion networks that incorporate different intermediate feature representation information in a deep neural network to alleviate the intrinsic sparsity of sketches  ...  In order to improve visual knowledge transfer from seen to unseen classes, we elaborate a coarse-to-fine conditional decoder that generates coarse-grained category-specific corresponding features first  ...  In the below table, multi-layer feature fusion, coarse-to-fine generation with multi-layer feature fusion, coarse-to-fine generation with multi-layer feature fusion and discriminability preserving are  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1904.01971v2">arXiv:1904.01971v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/uthgdpzqynb75bjcrlcrl5u744">fatcat:uthgdpzqynb75bjcrlcrl5u744</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200929202617/https://arxiv.org/pdf/1904.01971v1.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/9d/27/9d279583634afd8e280cb62bba17afc3402a4db0.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1904.01971v2" 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>

Multi-task Cascade Convolution Neural Networks for Automatic Thyroid Nodule Detection and Recognition

Wenfeng Song, Shuai Li, Ji Liu, Hong Qin, Bo Zhang, Zhang Shuyang, Aimin Hao
<span title="">2018</span> <i title="Institute of Electrical and Electronics Engineers (IEEE)"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/q2z26obphvchndqieqd65vltle" style="color: black;">IEEE journal of biomedical and health informatics</a> </i> &nbsp;
Within our framework, the potential regions of interest after initial detection are further fed to the spatial pyramid augmented CNNs to embed multi-scale discriminative information for finegrained thyroid  ...  It may be noted that, our framework is built upon a large number of clinically-confirmed thyroid ultrasound images with accurate and detailed ground truth labels.  ...  Different from the original SSD, we add multi-scale layers to fit for the thyroid nodules and arrive at coarse recognition. Multi-scale Detection Network.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/jbhi.2018.2852718">doi:10.1109/jbhi.2018.2852718</a> <a target="_blank" rel="external noopener" href="https://www.ncbi.nlm.nih.gov/pubmed/29994412">pmid:29994412</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/6lnvhu45a5fctl2vgymtwi5liy">fatcat:6lnvhu45a5fctl2vgymtwi5liy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20190223130619/http://pdfs.semanticscholar.org/501e/4b22eb8b5894a65cd14bb87d0206748d4eb2.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/50/1e/501e4b22eb8b5894a65cd14bb87d0206748d4eb2.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/jbhi.2018.2852718"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

Learning scale-variant and scale-invariant features for deep image classification

Nanne van Noord, Eric Postma
<span title="">2017</span> <i title="Elsevier BV"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/jm6w2xclfzguxnhmnmq5omebpi" style="color: black;">Pattern Recognition</a> </i> &nbsp;
Convolutional Neural Networks (CNNs) require large image corpora to be trained on classification tasks.  ...  We propose a multi-scale CNN method to encourage the recognition of both types of features and evaluate it on a challenging image classification task involving task-relevant characteristics at multiple  ...  Acknowledgement We would like to thank the anonymous reviewers for their insightful and constructive comments.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1016/j.patcog.2016.06.005">doi:10.1016/j.patcog.2016.06.005</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/yqfdq5yo5bdrtpn765mhmxolfe">fatcat:yqfdq5yo5bdrtpn765mhmxolfe</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20190419180911/https://core.ac.uk/download/pdf/82217181.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/fe/69/fe692a2dc651d0b53b617115e1d9121cb99dac82.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1016/j.patcog.2016.06.005"> <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>

Hierarchical Modified Fast R-CNN for Object Detection

Arindam Chaudhuri
<span title="2021-12-17">2021</span> <i title="Slovenian Association Informatika"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/w2vs4367uvgttdfo3ce3gi6rbe" style="color: black;">Informatica (Ljubljana, Tiskana izd.)</a> </i> &nbsp;
For large-scale recognition tasks, scalability is done considering conditional execution of fine category classifiers and layer parameters compression.  ...  In object detection there is high degree of skewedness for objects ' visual separability. It is difficult to distinguish object categories which demand dedicated classification.  ...  Input images are resized into different scales and multi-scale windows are used to slide through these images in order to capture information about multi-scale and different aspect ratios of objects.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.31449/inf.v45i7.3732">doi:10.31449/inf.v45i7.3732</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/rsvbpu2iq5abfkkk2nbt5box6m">fatcat:rsvbpu2iq5abfkkk2nbt5box6m</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220116212223/https://www.informatica.si/index.php/informatica/article/download/3732/1649" 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/6a/f0/6af0c85db171f9c412e12ea1e8a1604118c7fb47.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.31449/inf.v45i7.3732"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="unlock alternate icon" style="background-color: #fb971f;"></i> Publisher / doi.org </button> </a>

Dynamic Capacity Networks [article]

Amjad Almahairi, Nicolas Ballas, Tim Cooijmans, Yin Zheng, Hugo Larochelle, Aaron Courville
<span title="2016-05-22">2016</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We focus our empirical evaluation on the Cluttered MNIST and SVHN image datasets.  ...  The selection is made using a novel gradient-based attention mechanism, that efficiently identifies input regions for which the DCN's output is most sensitive and to which we should devote more capacity  ...  We would like to thank the developers of Theano (Bergstra et al., 2011; Bastien et al., 2012) and Blocks/Fuel (Van Merriënboer et al., 2015) for developing such powerful tools for scientific computing  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1511.07838v7">arXiv:1511.07838v7</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/tbqxtjtywzd4zfcvecstebfj6q">fatcat:tbqxtjtywzd4zfcvecstebfj6q</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200924193016/https://arxiv.org/pdf/1511.07838v7.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/9c/90/9c90531672d5aaf94c005608b12648fa7b5583dc.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1511.07838v7" 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 Novel Orthogonality Loss for Deep Hierarchical Multi-Task Learning

Guiqing He, Yincheng Huo, Mingyao He, Haixi Zhang, Jianping Fan
<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, a novel loss function is proposed to measure the correlation among different learning tasks and select useful feature components for each classification task.  ...  other two types of networks on image classification.  ...  The knowledge maps are widely used in large-scale classification task as it can efficiently organize large-scale object classes in a course to fine fashion.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/access.2020.2985991">doi:10.1109/access.2020.2985991</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/y5bvda5v2vcirnv4wbx5ztifum">fatcat:y5bvda5v2vcirnv4wbx5ztifum</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201107203157/https://ieeexplore.ieee.org/ielx7/6287639/8948470/09057591.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/17/e5/17e58061dea2ad94c76c2f0ddf52f66cc49160ef.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/access.2020.2985991"> <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>

Feature Selection based Hierarchical Deep Network for Image Classification

Guiqing He, Jiaqi Ji, Haixi Zhang, Yuelei Xu, Jianping Fan
<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;
First, the concept ontology is built for organizing large-scale image classes hierarchically in a coarse-to-fine fashion.  ...  The experimental results on three datasets show that adding a feature selection module in a hierarchical deep network can perform better performance in large-scale image classification.  ...  of natural world to large-scale image classification.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/access.2020.2966651">doi:10.1109/access.2020.2966651</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/5wkf36sly5h43ek4dyo5vmjmo4">fatcat:5wkf36sly5h43ek4dyo5vmjmo4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201107161806/https://ieeexplore.ieee.org/ielx7/6287639/8948470/08959205.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/25/d2/25d268e432ad399b1924c4914f8be38887bf3407.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/access.2020.2966651"> <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>

Look Closer to See Better: Recurrent Attention Convolutional Neural Network for Fine-Grained Image Recognition

Jianlong Fu, Heliang Zheng, Tao Mei
<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;
The proposed RA-CNN is optimized by an intra-scale classification loss and an inter-scale ranking loss, to mutually learn accurate region attention and fine-grained representation.  ...  The APN starts from full images, and iteratively generates region attention from coarse to fine by taking previous predictions as a reference, while a finer scale network takes as input an amplified attended  ...  Multi-scale Joint Representation Once the proposed RA-CNN has been trained at each scale, we can obtain multi-scale representations from fullsize images to multiple coarse-to-fine region attention.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/cvpr.2017.476">doi:10.1109/cvpr.2017.476</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/cvpr/FuZM17.html">dblp:conf/cvpr/FuZM17</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/xfv2yoyyqnehngfad6g22hmh2q">fatcat:xfv2yoyyqnehngfad6g22hmh2q</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200210154850/http://openaccess.thecvf.com/content_cvpr_2017/papers/Fu_Look_Closer_to_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/32/3a/323a3e4b1b8a838a1738e8538bfc15a2029c68af.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/cvpr.2017.476"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

Detecting 11K Classes: Large Scale Object Detection Without Fine-Grained Bounding Boxes

Hao Yang, Hao Wu, Hao Chen
<span title="">2019</span> <i title="IEEE"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/753trptklbb4nj6jquqadzwwdu" style="color: black;">2019 IEEE/CVF International Conference on Computer Vision (ICCV)</a> </i> &nbsp;
In this paper, we propose a semi-supervised large scale fine-grained detection method, which only needs bounding box annotations of a smaller number of coarsegrained classes and image-level labels of large  ...  However, these methods require fully annotated object bounding boxes for training, which are incredibly hard to scale up due to the high annotation cost.  ...  We believe a better way to train large-scale fine-grained detector is through semi-supervised learning with coarse-grained detection data and fine-grained classification data.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/iccv.2019.00990">doi:10.1109/iccv.2019.00990</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/iccv/YangWC19.html">dblp:conf/iccv/YangWC19</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/euiavla6uvfs3id5sobuul6v5y">fatcat:euiavla6uvfs3id5sobuul6v5y</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200709075507/https://assets.amazon.science/53/3e/b46992344be8bedff7f26d760108/detecting-11k-classes-large-sclae-object-detection-without-fine-grained-bounding-boxes.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/d0/f3/d0f30b0638a7badb1aad695fc180a663fad74f71.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/iccv.2019.00990"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

Multi-scale Recognition with DAG-CNNs

Songfan Yang, Deva Ramanan
<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;
We explore multi-scale convolutional neural nets (CNNs) for image classification. Contemporary approaches extract features from a single output layer.  ...  We use DAG-CNNs to learn a set of multiscale features that can be effectively shared between coarse and fine-grained classification tasks.  ...  We extract multi-scale features from multiple layers to simultaneously distinguish coarse and fine classes.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/iccv.2015.144">doi:10.1109/iccv.2015.144</a> <a target="_blank" rel="external noopener" href="https://dblp.org/rec/conf/iccv/YangR15.html">dblp:conf/iccv/YangR15</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/yixsrpsdtfhglbv5p3ehw7lthq">fatcat:yixsrpsdtfhglbv5p3ehw7lthq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20161020131509/http://vision.ics.uci.edu:80/papers/YangR_ICCV_2015/YangR_ICCV_2015.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/d5/aa/d5aaa0f136234da86cf56ff67e493c2f15cac920.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/iccv.2015.144"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>
&laquo; Previous Showing results 1 &mdash; 15 out of 13,868 results