{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,11]],"date-time":"2026-01-11T00:56:32Z","timestamp":1768092992952,"version":"3.49.0"},"reference-count":44,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"1","license":[{"start":{"date-parts":[[2022,1,1]],"date-time":"2022-01-01T00:00:00Z","timestamp":1640995200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2022,1,1]],"date-time":"2022-01-01T00:00:00Z","timestamp":1640995200000},"content-version":"am","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2022,1,1]],"date-time":"2022-01-01T00:00:00Z","timestamp":1640995200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2022,1,1]],"date-time":"2022-01-01T00:00:00Z","timestamp":1640995200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["1725447"],"award-info":[{"award-number":["1725447"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["1817037"],"award-info":[{"award-number":["1817037"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["1730309"],"award-info":[{"award-number":["1730309"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["61876215"],"award-info":[{"award-number":["61876215"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Beijing Academy of Artificial Intelligence"},{"name":"Institute for Guo Qiang"},{"DOI":"10.13039\/501100004147","name":"Tsinghua University","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100004147","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Key Scientific Technological Innovation Research"},{"name":"Zhejiang laboratory"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Pattern Anal. Mach. Intell."],"published-print":{"date-parts":[[2022,1,1]]},"DOI":"10.1109\/tpami.2020.3010201","type":"journal-article","created":{"date-parts":[[2020,7,17]],"date-time":"2020-07-17T20:44:41Z","timestamp":1595018681000},"page":"13-31","source":"Crossref","is-referenced-by-count":27,"title":["A Comprehensive and Modularized Statistical Framework for Gradient Norm Equality in Deep Neural Networks"],"prefix":"10.1109","volume":"44","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9601-4586","authenticated-orcid":false,"given":"Zhaodong","family":"Chen","sequence":"first","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5172-9411","authenticated-orcid":false,"given":"Lei","family":"Deng","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5240-7238","authenticated-orcid":false,"given":"Bangyan","family":"Wang","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8994-431X","authenticated-orcid":false,"given":"Guoqi","family":"Li","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2093-1788","authenticated-orcid":false,"given":"Yuan","family":"Xie","sequence":"additional","affiliation":[]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.123"},{"key":"ref2","article-title":"All you need is a good init","author":"Mishkin","year":"2016"},{"key":"ref3","first-page":"5389","article-title":"Dynamical isometry and a mean field theory of CNNs: How to train 10,000-layer vanilla convolutional neural networks","author":"Xiao","year":"2018"},{"key":"ref4","article-title":"Residual learning without normalization via better initialization","author":"Zhang","year":"2019"},{"key":"ref5","first-page":"448","article-title":"Batch normalization: Accelerating deep network training by reducing internal covariate shift","author":"Ioffe","year":"2015"},{"key":"ref6","first-page":"901","article-title":"Weight normalization: A simple reparameterization to accelerate training of deep neural networks","volume-title":"Proc. Advances Neural Inf. Process. Syst.","author":"Salimans"},{"key":"ref7","article-title":"Weight standardization","author":"Qiao","year":"2019"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.243"},{"key":"ref10","first-page":"3360","article-title":"Exponential expressivity in deep neural networks through transient chaos","volume-title":"Proc. Advances Neural Inf. Process. Syst.","author":"Poole"},{"key":"ref11","article-title":"A mean field theory of batch normalization","author":"Yang","year":"2019"},{"key":"ref12","article-title":"Deep information propagation","author":"Schoenholz","year":"2017"},{"key":"ref13","first-page":"4785","article-title":"Resurrecting the sigmoid in deep learning through dynamical isometry: theory and practice","volume-title":"Proc. Advances Neural Inf. Process. Syst.","author":"Pennington"},{"key":"ref14","first-page":"1924","article-title":"The emergence of spectral universality in deep networks","volume":"84","author":"Pennington","year":"2018"},{"key":"ref15","first-page":"2221","article-title":"Dynamical isometry is achieved in residual networks in a universal way for any activation function","author":"Tarnowski","year":"2019"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2019.2931991"},{"key":"ref17","article-title":"Gradients explode-deep networks are shallow-resnet explained","author":"Philipp","year":"2018"},{"key":"ref18","article-title":"The benefits of over-parameterization at initialization in deep ReLU networks","author":"Arpit","year":"2019"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1007\/978-1-4939-6942-5"},{"key":"ref20","article-title":"Partial freeness of random matrices","author":"Chen","year":"2012"},{"key":"ref21","first-page":"1168","article-title":"Normalization propagation: A parametric technique for removing internal covariate shift in deep networks","volume":"48","author":"Arpit","year":"2016"},{"key":"ref22","first-page":"971","article-title":"Self-normalizing neural networks","volume-title":"Proc. Advances Neural Inf. Process. Syst.","author":"Klambauer"},{"key":"ref23","article-title":"Mixup: Beyond empirical risk minimization","author":"Zhang","year":"2018"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2020.2978753"},{"key":"ref25","article-title":"Comparison of batch normalization and weight normalization algorithms for the large-scale image classification","author":"Gitman","year":"2017"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01261-8_1"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/EMC2.2018.00011"},{"key":"ref28","article-title":"Exact solutions to the nonlinear dynamics of learning in deep linear neural networks","author":"Saxe","year":"2014"},{"key":"ref29","article-title":"Non-hermitian random matrix theory for mimo channels","author":"Cakmak","year":"2012"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/5.726791"},{"key":"ref31","article-title":"Very deep convolutional networks for large-scale image recognition","author":"Simonyan","year":"2014"},{"key":"ref32","article-title":"Information geometry of orthogonal initializations and training","author":"Sok\u00f3","year":"2020"},{"key":"ref33","article-title":"Mean-field behaviour of neural tangent kernel for deep neural networks","author":"Hayou","year":"2019"},{"key":"ref34","article-title":"Mobilenets: Efficient convolutional neural networks for mobile vision applications","author":"Howard","year":"2017"},{"key":"ref35","first-page":"1097","article-title":"Imagenet classification with deep convolutional neural networks","volume-title":"Proc. Advances Neural Inf. Process. Syst.","author":"Krizhevsky"},{"key":"ref36","first-page":"315","article-title":"Deep sparse rectifier neural networks","volume-title":"Proc. 14th Int. Conf. Artif. Intell. Statist.","author":"Glorot"},{"key":"ref37","article-title":"Empirical evaluation of rectified activations in convolutional network","author":"Xu","year":"2015"},{"key":"ref38","article-title":"Instance normalization: The missing ingredient for fast stylization","author":"Ulyanov","year":"2016"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2018.2876179"},{"key":"ref40","article-title":"Towards understanding regularization in batch normalization","author":"Luo","year":"2019"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.4324\/9781410605337-29"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00907"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6420\/aa9a90"},{"key":"ref44","article-title":"Exact information propagation through fully-connected feed forward neural networks","author":"Burkholz","year":"2018"}],"container-title":["IEEE Transactions on Pattern Analysis and Machine Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/ieeexplore.ieee.org\/ielam\/34\/9639876\/9143512-aam.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/34\/9639876\/09143512.pdf?arnumber=9143512","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,1,9]],"date-time":"2024-01-09T23:43:47Z","timestamp":1704843827000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/9143512\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,1,1]]},"references-count":44,"journal-issue":{"issue":"1"},"URL":"https:\/\/doi.org\/10.1109\/tpami.2020.3010201","relation":{},"ISSN":["0162-8828","2160-9292","1939-3539"],"issn-type":[{"value":"0162-8828","type":"print"},{"value":"2160-9292","type":"electronic"},{"value":"1939-3539","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,1,1]]}}}