{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,2,21]],"date-time":"2025-02-21T20:22:00Z","timestamp":1740169320856,"version":"3.37.3"},"reference-count":55,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","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:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"funder":[{"DOI":"10.13039\/501100003725","name":"National Research Foundation (NRF) through Korean Government","doi-asserted-by":"publisher","award":["2021R1A2C3006659"],"award-info":[{"award-number":["2021R1A2C3006659"]}],"id":[{"id":"10.13039\/501100003725","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Institute of Information & Communications Technology Planning & Evaluation (IITP) through Korean Government","award":["2021-0-01343","2022-0-00953"],"award-info":[{"award-number":["2021-0-01343","2022-0-00953"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Access"],"published-print":{"date-parts":[[2022]]},"DOI":"10.1109\/access.2022.3231455","type":"journal-article","created":{"date-parts":[[2022,12,21]],"date-time":"2022-12-21T18:51:49Z","timestamp":1671648709000},"page":"133828-133841","source":"Crossref","is-referenced-by-count":1,"title":["Model Compression via Position-Based Scaled Gradient"],"prefix":"10.1109","volume":"10","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-1334-4649","authenticated-orcid":false,"given":"Jangho","family":"Kim","sequence":"first","affiliation":[{"name":"School of Artificial Intelligence, College of Computer Science, Kookmin University, Seoul, Republic of Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kiyoon","family":"Yoo","sequence":"additional","affiliation":[{"name":"Department of Intelligence and Information, GSCST, Seoul National University, Seoul, Republic of Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1792-0327","authenticated-orcid":false,"given":"Nojun","family":"Kwak","sequence":"additional","affiliation":[{"name":"Department of Intelligence and Information, GSCST, Seoul National University, Seoul, Republic of Korea"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.2307\/1267352"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1111\/j.2517-6161.1996.tb02080.x"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1503.02531"},{"key":"ref4","first-page":"2760","article-title":"Paraphrasing complex network: Network compression via factor transfer","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Kim"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1137\/0801001"},{"key":"ref6","first-page":"1","article-title":"Defensive quantization: When efficiency meets robustness","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Lin"},{"key":"ref7","first-page":"5767","article-title":"Improved training of Wasserstein GANs","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Gulrajani"},{"key":"ref8","first-page":"1","article-title":"Gradient \u21131 regularization for quantization robustness","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Alizadeh"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW50498.2020.00129"},{"key":"ref11","article-title":"Quantizing deep convolutional networks for efficient inference: A whitepaper","author":"Krishnamoorthi","year":"2018","journal-title":"arXiv:1806.08342"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00141"},{"key":"ref13","first-page":"7948","article-title":"Post training 4-bit quantization of convolutional networks for rapid-deployment","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Banner"},{"key":"ref14","first-page":"7543","article-title":"Improving neural network quantization without retraining using outlier channel splitting","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Zhao"},{"key":"ref15","first-page":"20415","article-title":"Position-based scaled gradient for model quantization and pruning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"33","author":"Kim"},{"key":"ref16","first-page":"10181","article-title":"The implicit and explicit regularization effects of dropout","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Wei"},{"key":"ref17","article-title":"Learned step size quantization","author":"Esser","year":"2019","journal-title":"arXiv:1902.08153"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00495"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW50498.2020.00356"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/ICCVW.2019.00363"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1007\/s10994-021-06053-z"},{"key":"ref22","article-title":"Up or down? Adaptive rounding for post-training quantization","author":"Nagel","year":"2020","journal-title":"arXiv:2004.10568"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v35i9.16950"},{"key":"ref24","article-title":"Dynamic collective intelligence learning: Finding efficient sparse model via refined gradients for pruned weights","author":"Kim","year":"2021","journal-title":"arXiv:2109.04660"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2021-248"},{"key":"ref26","first-page":"19637","article-title":"Only train once: A one-shot neural network training and pruning framework","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Chen"},{"key":"ref27","article-title":"The lottery ticket hypothesis: Finding sparse, trainable neural networks","author":"Frankle","year":"2018","journal-title":"arXiv:1803.03635"},{"key":"ref28","first-page":"9908","article-title":"Sparse training via boosting pruning plasticity with neuroregeneration","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Liu"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2021.11.029"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2021.3084856"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01270-0_19"},{"key":"ref32","article-title":"Pruning filters for efficient ConvNets","author":"Li","year":"2016","journal-title":"arXiv:1608.08710"},{"key":"ref33","article-title":"Comparing rewinding and fine-tuning in neural network pruning","author":"Renda","year":"2020","journal-title":"arXiv:2003.02389"},{"key":"ref34","article-title":"To prune, or not to prune: Exploring the efficacy of pruning for model compression","author":"Zhu","year":"2017","journal-title":"arXiv:1710.01878"},{"key":"ref35","first-page":"1","article-title":"Learning sparse neural networks through L0 regularization","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Louizos"},{"key":"ref36","first-page":"1","article-title":"SNIP: Single-shot network pruning based on connection sensitivity","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Lee"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.3390\/math10122126"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/ICPR48806.2021.9412615"},{"key":"ref39","first-page":"1","article-title":"Entropy-SGD: Biasing gradient descent into wide valleys","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Chaudhari"},{"key":"ref40","first-page":"192","article-title":"The loss surfaces of multilayer networks","volume-title":"Proc. Artif. Intell. Statist.","author":"Choromanska"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1017\/CBO9780511804441"},{"key":"ref42","article-title":"Adam: A method for stochastic optimization","author":"Kingma","year":"2014","journal-title":"arXiv:1412.6980"},{"key":"ref43","first-page":"1135","article-title":"Learning both weights and connections for efficient neural network","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Han"},{"key":"ref44","first-page":"1","article-title":"Dynamic model pruning with feedback","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Lin"},{"key":"ref45","article-title":"Sparse networks from scratch: Faster training without losing performance","author":"Dettmers","year":"2019","journal-title":"arXiv:1907.04840"},{"key":"ref46","first-page":"4646","article-title":"Parameter efficient training of deep convolutional neural networks by dynamic sparse reparameterization","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Mostafa"},{"key":"ref47","first-page":"8557","article-title":"AC\/DC: Alternating compressed\/decompressed training of deep neural networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Peste"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01237-3_23"},{"key":"ref49","article-title":"The state of sparsity in deep neural networks","author":"Gale","year":"2019","journal-title":"arXiv:1902.09574"},{"key":"ref50","article-title":"Very deep convolutional networks for large-scale image recognition","author":"Simonyan","year":"2014","journal-title":"arXiv:1409.1556"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.243"},{"key":"ref52","article-title":"QKD: Quantization-aware knowledge distillation","author":"Kim","year":"2019","journal-title":"arXiv:1911.12491"},{"volume-title":"Deep Learning","year":"2016","author":"Goodfellow","key":"ref53"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1088\/1742-5468\/ab39d9"},{"key":"ref55","first-page":"6389","article-title":"Visualizing the loss landscape of neural nets","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Li"}],"container-title":["IEEE Access"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/6287639\/9668973\/09996406.pdf?arnumber=9996406","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,3,2]],"date-time":"2024-03-02T11:35:35Z","timestamp":1709379335000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/9996406\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022]]},"references-count":55,"URL":"https:\/\/doi.org\/10.1109\/access.2022.3231455","relation":{},"ISSN":["2169-3536"],"issn-type":[{"type":"electronic","value":"2169-3536"}],"subject":[],"published":{"date-parts":[[2022]]}}}