Filters








898 Hits in 4.1 sec

SpectralFormer: Rethinking Hyperspectral Image Classification with Transformers [article]

Danfeng Hong and Zhu Han and Jing Yao and Lianru Gao and Bing Zhang and Antonio Plaza and Jocelyn Chanussot
<span title="2021-11-20">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
To solve this issue, we rethink HS image classification from a sequential perspective with transformers, and propose a novel backbone network called SpectralFormer.  ...  Beyond band-wise representations in classic transformers, SpectralFormer is capable of learning spectrally local sequence information from neighboring bands of HS images, yielding group-wise spectral embeddings  ...  University of Houston, with extensive the use of multi-head attention, where multiple self-attention ablation studies.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2107.02988v2">arXiv:2107.02988v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/iw67o2iwhjafbhhrwogcswyk7u">fatcat:iw67o2iwhjafbhhrwogcswyk7u</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210715032311/https://arxiv.org/pdf/2107.02988v1.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/dc/6a/dc6a586de0b0d838edd34f2e427fbaaa12f40f40.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2107.02988v2" 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>

Scalable Graph Convolutional Networks with Fast Localized Spectral Filter for Directed Graphs

Chensheng Li, Xiaowei Qin, Xiaodong Xu, Dujia Yang, Guo Wei
<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;
Compared with spatial-based GCNNs, spectral-based GCNNs are capable of highly exploiting graph structure information, but always regard graphs undiredcted.  ...  FDGCN can directly work on directed graphs and can scale to large graphs as the convolution operation is linear with the number of edges.  ...  With this expansion of K th -order Chebyshev polynomials in Laplacian, a spectral graph filter can be performed without Fourier transform, which implies no eigen-decomposition is required.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/access.2020.2999520">doi:10.1109/access.2020.2999520</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/eylcctoh2ffynibbmj3mt3tpvy">fatcat:eylcctoh2ffynibbmj3mt3tpvy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210429114236/https://ieeexplore.ieee.org/ielx7/6287639/8948470/09106312.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/5e/60/5e607d3796a20776095f0cfd92ab49cc91fbf0bc.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/access.2020.2999520"> <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>

Network representation learning: A macro and micro view

Xueyi Liu, Jie Tang
<span title="">2021</span> <i title="Elsevier BV"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/bm3baoxewvempaqpsm7ytjgkci" style="color: black;">AI Open</a> </i> &nbsp;
Graph is a universe data structure that is widely used to organize data in real-world.  ...  Various real-word networks like the transportation network, social and academic network can be represented by graphs.  ...  Universal Graph Spectral Filters. Graph spectral filters have close connections with spatial properties.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1016/j.aiopen.2021.02.001">doi:10.1016/j.aiopen.2021.02.001</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/6ktfheijvjdnfhja5oqobse5b4">fatcat:6ktfheijvjdnfhja5oqobse5b4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210728192717/https://pdf.sciencedirectassets.com/777606/1-s2.0-S2666651021X00022/1-s2.0-S2666651021000024/main.pdf?X-Amz-Security-Token=IQoJb3JpZ2luX2VjEHMaCXVzLWVhc3QtMSJHMEUCIQDyrYdUnQpOTBS40lRikdkxYyPMvxxJGTSBQvHEvCzD7gIgVrMCvCmTyZHQ%2BjeO8h6XPdiXvnQ2ukRlUmxlBBnn64kq%2BgMIfBAEGgwwNTkwMDM1NDY4NjUiDLaY%2BmwVMYrYPq2AiSrXAwK0%2F9PxOxo09NATOre%2B0mF%2FU2xdvajTiBvbCP3hxTamJF2S7WTrAm%2FX9adfaktTY8C5SCW2qeELtLxzYQ0nfSoBTuzlGAvcGFTdrpbjik9meuMh0olbzfwRurK8fXyBxdW4Xtrk1nzSsYqu%2FYP6VV80wCg%2Bfcx8qyo5JCnjt5OFuFxJH4imuAEi0KI8PxqsiCK6DbL%2BL2qSzVEW0M0x7XvIAzK5ECG%2FmZVscsty34PGDFwGpMb3FPJ9%2FE%2FL5kicYTlPomhTJRBPpmsSutd2uKPnEErZyHyKNO4PQ3eESoVh6aNwb8wCXhnaa3iMtCAhCyzBODzUGX6wbSZ5bxfPUKejOHc%2BS4pv%2F3IyLWBEL4COQsxeU5Dc5a46qnWx2VbXYhABEPw8akYn7JsU6vnW0Q8SbFREYF%2BYjG2%2BuMYQgAtS0gqGx4QpVHlxXygw5Q18B5skcVquYIArqar71PJioW7zoQXaK2LWIqTNTMDFpV4iZ2n9%2FmponOkDvHzVcKfeNqNqDAD72UkcPqB71snjBnU2xpdiSGSwrVM1w6tyi6DrF9R3l4a27VSDjGR1v8TR8d9z4oFPesupfE2fGA%2FFAzxqyUEOksN6opp6AJHx34ghztXEtdjHQDCJwYaIBjqlAVARTVp1qASvBVovvvKh7iE6MFeRk4TU9NYFVJ%2FnjExEtbJD%2B7tNZNoX58NdBFjwGS6MyDHYxaBP99SfBHeA3TFZS5Dn1AQMyiVjXVQJu6xyCgb8Jf9ngk06G931%2FFsLtSjZOWcn8owNwSnQRbDfogNd%2BnmjDK1dJkflLD7t4k9KZhjRDW%2BNKJo2rUCJCQQ%2BZ%2B9UBnGuA2FMz01PV1M87MYNgEEjbw%3D%3D&amp;X-Amz-Algorithm=AWS4-HMAC-SHA256&amp;X-Amz-Date=20210728T192712Z&amp;X-Amz-SignedHeaders=host&amp;X-Amz-Expires=300&amp;X-Amz-Credential=ASIAQ3PHCVTYSPSW2OG5%2F20210728%2Fus-east-1%2Fs3%2Faws4_request&amp;X-Amz-Signature=3ea51c3f6dddd22b4d29778f2a860f706d9a6b88421e796ae0097f3efb1be764&amp;hash=2191c059ab7cd5e3f6afe22008a808d5092911a5e26c9366d4eee952686eb5db&amp;host=68042c943591013ac2b2430a89b270f6af2c76d8dfd086a07176afe7c76c2c61&amp;pii=S2666651021000024&amp;tid=spdf-492fa2ae-a868-4a43-aa47-d011fd1a0352&amp;sid=23e388911c8e294fde399e453b0555f59d51gxrqa&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/d6/78/d678a4233980b22907709269f8be2a070ae2d792.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1016/j.aiopen.2021.02.001"> <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>

Interferometric Graph Transform: a Deep Unsupervised Graph Representation [article]

Edouard Oyallon
<span title="2020-06-10">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Our first contribution is to propose a generic, complex-valued spectral graph architecture obtained from a generalization of the Euclidean Fourier transform.  ...  We propose the Interferometric Graph Transform (IGT), which is a new class of deep unsupervised graph convolutional neural network for building graph representations.  ...  Rethinking knowledge graph propaga- Oyallon, E., Belilovsky, E., and Zagoruyko, S. Scaling the tion for zero-shot learning.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2006.05722v1">arXiv:2006.05722v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/dupcll67t5dffaxfiq54nft4wq">fatcat:dupcll67t5dffaxfiq54nft4wq</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200829134613/https://arxiv.org/pdf/2006.05722v1.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/ad/77/ad770d6efc11c34e410bdcda4e1a380c35006bfe.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2006.05722v1" 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>

2021 Index IEEE Transactions on Image Processing Vol. 30

<span title="">2021</span> <i title="Institute of Electrical and Electronics Engineers (IEEE)"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/dhlhr4jqkbcmdbua2ca45o7kru" style="color: black;">IEEE Transactions on Image Processing</a> </i> &nbsp;
Ruan, D., +, TIP 2021 6906-6916 SRGAT: Single Image Super-Resolution With Graph Attention Network.  ...  Chen, C., +, TIP 2021 3995-4007 Rethinking Motion Representation: Residual Frames With 3D ConvNets.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/tip.2022.3142569">doi:10.1109/tip.2022.3142569</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/z26yhwuecbgrnb2czhwjlf73qu">fatcat:z26yhwuecbgrnb2czhwjlf73qu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220114082313/https://ieeexplore.ieee.org/ielx7/83/9263394/09679464.pdf?tp=&amp;arnumber=9679464&amp;isnumber=9263394&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/88/c2/88c257e9b04f04be0d262c7cb19aa9d2a764c5f3.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/tip.2022.3142569"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

Network representation learning: A macro and micro view [article]

Xueyi Liu, Jie Tang
<span title="2021-11-21">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Graph is a universe data structure that is widely used to organize data in real-world.  ...  Various real-word networks like the transportation network, social and academic network can be represented by graphs.  ...  Universal Graph Spectral Filters. Graph spectral filters have close connections with spatial properties.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2111.10772v1">arXiv:2111.10772v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/rmmplc4qbzhkxloauzz3micbay">fatcat:rmmplc4qbzhkxloauzz3micbay</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211201174406/https://arxiv.org/ftp/arxiv/papers/2111/2111.10772.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/0f/54/0f540c893964e2ccc257a12878ed12f397582113.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2111.10772v1" 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>

Bridging the Gap between Spatial and Spectral Domains: A Unified Framework for Graph Neural Networks [article]

Zhiqian Chen, Fanglan Chen, Lei Zhang, Taoran Ji, Kaiqun Fu, Liang Zhao, Feng Chen, Lingfei Wu, Charu Aggarwal, Chang-Tien Lu
<span title="2021-08-07">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Graph neural networks (GNNs) are designed to deal with graph-structural data that classical deep learning does not easily manage.  ...  Prior research has primarily concentrated on categorizing existing models, with little attention paid to their intrinsic connections.  ...  Graph Attention Model (GAT) .  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2107.10234v3">arXiv:2107.10234v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/5pu74kvwf5hnzmsch6hewiydzm">fatcat:5pu74kvwf5hnzmsch6hewiydzm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210812055247/https://arxiv.org/pdf/2107.10234v3.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/34/6c/346c30f1c2e914a93d71165911c6be3752b2cef0.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2107.10234v3" 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>

Neighbor Enhanced Graph Convolutional Networks for Node Classification and Recommendation [article]

Hao Chen, Zhong Huang, Yue Xu, Zengde Deng, Feiran Huang, Peng He, Zhoujun Li
<span title="2022-03-30">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
The recently proposed Graph Convolutional Networks (GCNs) have achieved significantly superior performance on various graph-related tasks, such as node classification and recommendation.  ...  recursively aggregate the information from all the neighbors or randomly sampled neighbor subsets, without explicitly identifying whether the aggregated neighbors provide useful information during the graph  ...  Dual Graph Attention Networks [55] instead utilize the multi-arm bandit to explore the user interests with graph attention.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2203.16097v1">arXiv:2203.16097v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/lgopkam345fipp76zsmowaylci">fatcat:lgopkam345fipp76zsmowaylci</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220520145755/https://arxiv.org/pdf/2203.16097v1.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/4d/81/4d811b2003443aa3c6c372de2f10af1947ad950f.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2203.16097v1" 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>

Table of contents

<span title="">2021</span> <i title="Institute of Electrical and Electronics Engineers (IEEE)"> <a target="_blank" rel="noopener" href="https://fatcat.wiki/container/4odsbtjobjalfki6xxabjpdu6y" style="color: black;">IEEE Transactions on Geoscience and Remote Sensing</a> </i> &nbsp;
Unal 3486 Fast Georeferenced Aerial Image Stitching With Absolute Rotation Averaging and Planar-Restricted Pose Graph .... ........................................................................ Y.  ...  Zhang 3473 Rethinking CNN-(Contents Continued on Page 2702) (Contents Continued from Page 2701)  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/tgrs.2021.3063896">doi:10.1109/tgrs.2021.3063896</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/miqrb4or7jbujhr2aqlvhxq3oi">fatcat:miqrb4or7jbujhr2aqlvhxq3oi</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210402220949/https://ieeexplore.ieee.org/ielx7/36/9386007/09386018.pdf?tp=&amp;arnumber=9386018&amp;isnumber=9386007&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/25/fc/25fca475c1757cd7708532d066b26d3d56bfcc3b.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.1109/tgrs.2021.3063896"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> ieee.com </button> </a>

Universal Deep GNNs: Rethinking Residual Connection in GNNs from a Path Decomposition Perspective for Preventing the Over-smoothing [article]

Jie Chen, Weiqi Liu, Zhizhong Huang, Junbin Gao, Junping Zhang, Jian Pu
<span title="2022-05-30">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In this paper, we investigate the forward and backward behavior of GNNs with residual connections from a novel path decomposition perspective.  ...  Entangled propagation and weight matrices cause gradient smoothing and prevent GNNs with residual connections from optimizing to the identity mapping.  ...  Graph Neural Networks The general GNN is composed of information aggregation and feature transformation [CN, 13, 39, 44] .  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2205.15127v1">arXiv:2205.15127v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/sqxvn3iyenb3pgxi7aa4gyplie">fatcat:sqxvn3iyenb3pgxi7aa4gyplie</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220605172442/https://arxiv.org/pdf/2205.15127v1.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/31/af/31af437aee1630bfdf2ae38ed459b4fe9d6150e8.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2205.15127v1" 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>

Beyond Low-frequency Information in Graph Convolutional Networks [article]

Deyu Bo and Xiao Wang and Chuan Shi and Huawei Shen
<span title="2021-01-04">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We tackle this challenge and propose a novel Frequency Adaptation Graph Convolutional Networks (FAGCN) with a self-gating mechanism, which can adaptively integrate different signals in the process of message  ...  Graph neural networks (GNNs) have been proven to be effective in various network-related tasks.  ...  Given a signal x ∈ R n , the graph Fourier transform is defined asx = U x, and the inverse graph Fourier transform is x = Ux.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2101.00797v1">arXiv:2101.00797v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/vzjdlp6yj5hmrfooxnzxm7v2z4">fatcat:vzjdlp6yj5hmrfooxnzxm7v2z4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210225200530/https://arxiv.org/pdf/2101.00797v1.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/fb/c1/fbc136c8c81cd89206dc0fcb54e16bd98df83b62.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2101.00797v1" 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>

Rethinking pooling in graph neural networks [article]

Diego Mesquita, Amauri H. Souza, Samuel Kaski
<span title="2020-10-22">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Graph pooling is a central component of a myriad of graph neural network (GNN) architectures.  ...  As an inheritance from traditional CNNs, most approaches formulate graph pooling as a cluster assignment problem, extending the idea of local patches in regular grids to graphs.  ...  We found that this basic convolution produces better results than the random walk with restart (RWR) and graph attention networks as suggested in [21] .  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2010.11418v1">arXiv:2010.11418v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/xezsvz4rqvd5lmvufrb2pc73f4">fatcat:xezsvz4rqvd5lmvufrb2pc73f4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20201024100436/https://arxiv.org/pdf/2010.11418v1.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/e0/b3/e0b34b9dd924768fce9b52bc59ffb3975deb0d63.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2010.11418v1" 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>

CDGNet: A Cross-Time Dynamic Graph-based Deep Learning Model for Traffic Forecasting [article]

Yuchen Fang, Yanjun Qin, Haiyong Luo, Fang Zhao, Liang Zeng, Bo Hui, Chenxing Wang
<span title="2021-12-06">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Prior methods use the pre-defined or learnable static graph to extract spatial correlations. However, the static graph-based methods fail to mine the evolution of the traffic network.  ...  In this paper, we propose a novel cross-time dynamic graph-based deep learning model, named CDGNet, for traffic forecasting.  ...  Rethinking Attention with Performers. In 9th International Con- model, but too large feature dimension leads to over-fitting.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2112.02736v1">arXiv:2112.02736v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/qspala2jy5ajbeuxndgonaka7i">fatcat:qspala2jy5ajbeuxndgonaka7i</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211208111107/https://arxiv.org/pdf/2112.02736v1.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/78/fb/78fbfb6096a2d53e9b89e882c562bf3eabd7915a.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2112.02736v1" 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>

Towards Efficient Single Image Dehazing and Desnowing [article]

Tian Ye and Sixiang Chen and Yun Liu and Erkang Chen and Yuche Li
<span title="2022-04-19">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Specifically, it adopts a lightweight Adaptive Gated Neural Network to estimate gated attention maps of the input image, while different task-specific experts with the same topology are jointly dispatched  ...  Although the current recovery algorithms targeting a specific condition have made impressive progress, it is not flexible enough to deal with various degradation types.  ...  kernel size 3 × 3 (VC), the Localbranch (LB), the Global-branch (GB), the Multi-scale Spectral Transform Block without Spectral Transform operation (MSTBwoST) and the Multi-branch Spectral Transform Block  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2204.08899v1">arXiv:2204.08899v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/rbd2lugkbnhevldj2h6dedfohy">fatcat:rbd2lugkbnhevldj2h6dedfohy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220517054937/https://arxiv.org/pdf/2204.08899v1.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/46/c3/46c305b71d6bdded5a6dfe9727ef756f9021f18d.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2204.08899v1" 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>

Weisfeiler and Leman Go Infinite: Spectral and Combinatorial Pre-Colorings [article]

Or Feldman, Amit Boyarski, Shai Feldman, Dani Kogan, Avi Mendelson, Chaim Baskin
<span title="2022-03-02">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Graph isomorphism testing is usually approached via the comparison of graph invariants.  ...  ., obtained via the Weisfeiler-Leman (WL) test) and spectral invariants.  ...  Then it encodes them into node features using a transformer with self-attention.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2201.13410v2">arXiv:2201.13410v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/nqvzk544szd4zhpdtikavrkj3a">fatcat:nqvzk544szd4zhpdtikavrkj3a</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220519143814/https://arxiv.org/pdf/2201.13410v2.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/21/1a/211af952734e053c01cae09f8f0aaca3b1564bf1.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2201.13410v2" 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 898 results