{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,9]],"date-time":"2026-05-09T00:16:08Z","timestamp":1778285768052,"version":"3.51.4"},"reference-count":34,"publisher":"Institution of Engineering and Technology (IET)","issue":"1","license":[{"start":{"date-parts":[[2025,10,17]],"date-time":"2025-10-17T00:00:00Z","timestamp":1760659200000},"content-version":"vor","delay-in-days":289,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"},{"start":{"date-parts":[[2025,1,1]],"date-time":"2025-01-01T00:00:00Z","timestamp":1735689600000},"content-version":"tdm","delay-in-days":0,"URL":"http:\/\/doi.wiley.com\/10.1002\/tdm_license_1.1"}],"funder":[{"DOI":"10.13039\/501100012389","name":"National Institute of Information and Communications Technology","doi-asserted-by":"publisher","award":["JPJ012368C06801"],"award-info":[{"award-number":["JPJ012368C06801"]}],"id":[{"id":"10.13039\/501100012389","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100007884","name":"Hoso Bunka Foundation","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100007884","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["ietresearch.onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["IET Image Processing"],"published-print":{"date-parts":[[2025,1]]},"abstract":"<jats:title>ABSTRACT<\/jats:title>\n                  <jats:p>Learned image compression (LIC) has become more and more important in recent years. The hyperprior\u2010module\u2010based LIC models, which use hyperprior module to predict the distribution of image features and improve entropy coder performance, have achieved remarkable rate\u2010distortion (RD) performance. However, the storage and memory costs of these LIC models are too high, resulting in higher difficulty to be applied to various devices, especially portable or edge devices. The storage and memory cost are directly linked to the parameter number. As a preliminary experiment, we manually assigned half channels for the hyperprior module in LIC models, reducing about 30% parameters in the model. The pruned models still kept similar RD performance to the original ones. This reveals that the hyperprior module in LIC models is highly redundant. In the meanwhile, LIC models with different reconstruction qualities require different amounts of parameters for the hyperprior module. Based on these phenomena, we propose a quality\u2010aware hyperprior pruning method that efficiently reduces the storage and memory cost of the hyperprior module and various context models. It consists of two parts. The first part is the pruning method itself, called enhanced ResRep on hyper path (ERHP). The second part is a quality\u2010aware threshold searching method, called pruning threshold searching (PTS), which prunes the hyperprior module based on the reconstruction qualities of LIC models. The experiments on various LIC models show that our methods reduce large volumes of storage cost (up to 74.6%) and memory cost (up to 41.5%), while keeping the performance the same before\u00a0pruning.<\/jats:p>","DOI":"10.1049\/ipr2.70231","type":"journal-article","created":{"date-parts":[[2025,10,17]],"date-time":"2025-10-17T10:21:35Z","timestamp":1760696495000},"update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Storage\u2010and\u2010Memory\u2010Efficient Learned Image Compression With Quality\u2010Aware Hyperprior Pruning"],"prefix":"10.1049","volume":"19","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9919-3022","authenticated-orcid":false,"given":"Ao","family":"Luo","sequence":"first","affiliation":[{"name":"XR Division KDDI Research, Inc.  Fujimino\u2010shi Saitama Japan"},{"name":"Department of Computer Science and Communication Engineering Waseda University Shinjuku\u2010ku Tokyo Japan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Diego","family":"Fujii","sequence":"additional","affiliation":[{"name":"XR Division KDDI Research, Inc.  Fujimino\u2010shi Saitama Japan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Keisuke","family":"Nonaka","sequence":"additional","affiliation":[{"name":"XR Division KDDI Research, Inc.  Fujimino\u2010shi Saitama Japan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Heming","family":"Sun","sequence":"additional","affiliation":[{"name":"Faculty of Engineering Yokohama National University Yokohama Japan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jiro","family":"Katto","sequence":"additional","affiliation":[{"name":"Department of Computer Science and Communication Engineering Waseda University Shinjuku\u2010ku Tokyo Japan"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"265","published-online":{"date-parts":[[2025,10,17]]},"reference":[{"key":"e_1_2_12_2_1","doi-asserted-by":"crossref","unstructured":"G. K.Wallace \u201cThe JPEG Still Picture Compression Standard \u201dIEEE Transactions on Consumer Electronics38 no.1(1992): xviii\u2013xxxiv.","DOI":"10.1109\/30.125072"},{"key":"e_1_2_12_3_1","doi-asserted-by":"publisher","DOI":"10.1117\/1.1469618"},{"key":"e_1_2_12_4_1","unstructured":"F.Bellard \u201cBpg Image Format \u201d accessed April 21 2018 https:\/\/bellard.org\/bpg(2015)."},{"key":"e_1_2_12_5_1","unstructured":"B.Bross J.Chen S.Liu andY.\u2010K.Wang \u201cJVET\u2010S2001 Versatile Video Coding (Draft 10) \u201d inJoint Video Exploration Team (JVET) of ITU\u2010T SG 16 WP 3 and ISO\/IEC JTC 1\/SC 29\/WG 11(ITU\/ISO\/IEC 2020)."},{"key":"e_1_2_12_6_1","unstructured":"J.Ball\u00e9 D.Minnen S.Singh S. J.Hwang andN.Johnston \u201cVariational Image Compression With a Scale Hyperprior \u201d inInternational Conference on Learning Representations (ICLR)(2018) Vancouver BC Canada April 30\u2013May 3 2018."},{"key":"e_1_2_12_7_1","unstructured":"D.Minnen J.Ball\u00e9 andG. D.Toderici \u201cJoint Autoregressive and Hierarchical Priors for Learned Image Compression \u201d inAdvances in Neural Information Processing Systems (NeurIPS)31(2018):10794\u201310803."},{"key":"e_1_2_12_8_1","doi-asserted-by":"crossref","unstructured":"Z.Cheng H.Sun M.Takeuchi andJ.Katto \u201cLearned Image Compression With Discretized Gaussian Mixture Likelihoods and Attention Modules \u201d inIEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)(IEEE 2020) 7939\u20137948.","DOI":"10.1109\/CVPR42600.2020.00796"},{"key":"e_1_2_12_9_1","doi-asserted-by":"crossref","unstructured":"D.MinnenandS.Singh \u201cChannel\u2010Wise Autoregressive Entropy Models for Learned Image Compression \u201d inIEEE International Conference on Image Processing (ICIP)(IEEE 2020) 3339\u20133343.","DOI":"10.1109\/ICIP40778.2020.9190935"},{"key":"e_1_2_12_10_1","unstructured":"Y.Zhu Y.Yang andT.Cohen \u201cTransformer\u2010Based Transform Coding \u201dposter presented at the International Conference on Learning Representations (ICLR 2022) virtual April 25\u201329 2022."},{"key":"e_1_2_12_11_1","doi-asserted-by":"crossref","unstructured":"R.Zou C.Song andZ.Zhang \u201cThe Devil is in the Details: Window\u2010Based Attention for Image Compression \u201d inProceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)(IEEE 2022) 17492\u201317501.","DOI":"10.1109\/CVPR52688.2022.01697"},{"key":"e_1_2_12_12_1","doi-asserted-by":"crossref","unstructured":"J.Liu H.Sun andJ.Katto \u201cLearned Image Compression With Mixed Transformer\u2010CNN Architectures \u201d inIEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)(IEEE 2023) 14388\u201314397.","DOI":"10.1109\/CVPR52729.2023.01383"},{"key":"e_1_2_12_13_1","unstructured":"\u201cThe Kodak Photocd Dataset \u201dKodak Lossless True Color Image Suite (Eastman Kodak Company 2013) http:\/\/r0k.us\/graphics\/kodak\/."},{"key":"e_1_2_12_14_1","doi-asserted-by":"publisher","DOI":"10.1145\/3474085.3475667"},{"key":"e_1_2_12_15_1","unstructured":"N.Johnston E.Eban A.Gordon andJ.Ball\u00e9 \u201cComputationally Efficient Neural Image Compression \u201dpreprint arXiv December 18 2019 https:\/\/arxiv.org\/abs\/1912.08771."},{"key":"e_1_2_12_16_1","doi-asserted-by":"crossref","unstructured":"T. A.MunnaandJ.Ascenso \u201cComplexity Scalable Learning\u2010Based Image Decoding \u201d inProceedings of the IEEE International Conference on Image Processing (ICIP)(IEEE 2023) 1860\u20131864.","DOI":"10.1109\/ICIP49359.2023.10222047"},{"key":"e_1_2_12_17_1","doi-asserted-by":"crossref","unstructured":"A.Luo H.Sun J.Liu andJ.Katto \u201cMemory\u2010Efficient Learned Image Compression With Pruned Hyperprior Module \u201d inIEEE International Conference on Image Processing (ICIP)(IEEE 2022) 3061\u20133065.","DOI":"10.1109\/ICIP46576.2022.9897854"},{"key":"e_1_2_12_18_1","doi-asserted-by":"crossref","unstructured":"A.Luo H.Sun J.Liu F.Lin andJ.Katto \u201cPTS\u2010LIC: Pruning Threshold Searching for Lightweight Learned Image Compression \u201d inIEEE International Conference on Visual Communications and Image Processing (VCIP)(IEEE 2023) 1\u20135.","DOI":"10.1109\/VCIP59821.2023.10402683"},{"key":"e_1_2_12_19_1","doi-asserted-by":"crossref","unstructured":"X.Ding T.Hao J.Tan et\u00a0al. \u201cResrep: Lossless CNN Pruning via Decoupling Remembering and Forgetting \u201d inIEEE\/CVF International Conference on Computer Vision (CVPR)(IEEE 2021) 4510\u20134520.","DOI":"10.1109\/ICCV48922.2021.00447"},{"key":"e_1_2_12_20_1","doi-asserted-by":"crossref","unstructured":"W.Shi J.Caballero F.Husz\u00e1r et\u00a0al. \u201cReal\u2010Time Single Image and Video Super\u2010Resolution Using an Efficient Sub\u2010Pixel Convolutional Neural Network \u201d inIEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)(IEEE 2016) 1874\u20131883.","DOI":"10.1109\/CVPR.2016.207"},{"key":"e_1_2_12_21_1","doi-asserted-by":"crossref","unstructured":"J.Ball\u00e9 V.Laparra andE. P.Simoncelli \u201cEnd\u2010to\u2010End Optimization of Nonlinear Transform Codes for Perceptual Quality \u201d inProceedings of the Picture Coding Symposium (PCS)(IEEE 2016) 1\u20135.","DOI":"10.1109\/PCS.2016.7906310"},{"key":"e_1_2_12_22_1","unstructured":"J.Ball\u00e9 V.Laparra andE. P.Simoncelli \u201cEnd\u2010to\u2010End Optimized Image Compression \u201dpaper presented at the 5th International Conference on Learning Representations (ICLR) Toulon France April 24\u201326 2017."},{"key":"e_1_2_12_23_1","doi-asserted-by":"crossref","unstructured":"Y.He X.Zhang andJ.Sun \u201cChannel Pruning for Accelerating Very Deep Neural Networks \u201d inIEEE International Conference on Computer Vision (ICCV)(IEEE 2017) 1389\u20131397.","DOI":"10.1109\/ICCV.2017.155"},{"key":"e_1_2_12_24_1","doi-asserted-by":"crossref","unstructured":"Z.Zhan Y.Gong P.Zhao et\u00a0al. \u201cAchieving on\u2010Mobile Real\u2010Time Super\u2010Resolution With Neural Architecture and Pruning Search \u201d inIEEE\/CVF International Conference on Computer Vision (ICCV)(IEEE 2021) 4821\u20134831.","DOI":"10.1109\/ICCV48922.2021.00478"},{"key":"e_1_2_12_25_1","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-9868.2005.00532.x"},{"key":"e_1_2_12_26_1","doi-asserted-by":"crossref","unstructured":"F.Yang L.Herranz Y.Cheng andM. G.Mozerov \u201cSlimmable Compressive Autoencoders for Practical Neural Image Compression \u201d inIEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)(IEEE 2021) 4998\u20135007.","DOI":"10.1109\/CVPR46437.2021.00496"},{"key":"e_1_2_12_27_1","doi-asserted-by":"crossref","unstructured":"M. A. F.HossainandF.Zhu \u201cStructured Pruning and Quantization for Learned Image Compression \u201d inIEEE International Conference on Image Processing (ICIP)(IEEE 2024) 3730\u20133736.","DOI":"10.1109\/ICIP51287.2024.10648236"},{"issue":"63","key":"e_1_2_12_28_1","first-page":"1","article-title":"Nearly\u2010Tight vc\u2010Dimension and Pseudodimension Bounds for Piecewise Linear Neural Networks","volume":"20","author":"Bartlett P. L.","year":"2019","journal-title":"Journal of Machine Learning Research"},{"key":"e_1_2_12_29_1","doi-asserted-by":"crossref","unstructured":"K.He X.Zhang S.Ren andJ.Sun \u201cDeep Residual Learning for Image Recognition \u201d inIEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)(IEEE 2016) 770\u2013778.","DOI":"10.1109\/CVPR.2016.90"},{"key":"e_1_2_12_30_1","first-page":"5998","article-title":"Attention is all you Need","volume":"30","author":"Vaswani A.","year":"2017","journal-title":"Advances in Neural Information Processing Systems (NeurIPS)"},{"key":"e_1_2_12_31_1","doi-asserted-by":"crossref","unstructured":"C.Schumann S.Ricco U.Prabhu V.Ferrari andC.Pantofaru \u201cA Step Toward More Inclusive People Annotations for Fairness \u201d inProceedings of the 2021 AAAI\/ACM Conference on AI Ethics and Society(ACM 2021) 916\u2013925.","DOI":"10.1145\/3461702.3462594"},{"key":"e_1_2_12_32_1","unstructured":"J.B\u00e9gaint F.Racap\u00e9 S.Feltman andA.Pushparaja \u201cCompressai: A Pytorch Library and Evaluation Platform for end\u2010to\u2010end Compression Research \u201dpreprint arXiv November 6 2020 https:\/\/arxiv.org\/abs\/2011.03029."},{"key":"e_1_2_12_33_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.image.2014.12.007"},{"key":"e_1_2_12_34_1","doi-asserted-by":"crossref","unstructured":"M. A. F.HossainandF.Zhu \u201cStructured Pruning and Quantization for Learned Image Compression \u201d in2024 IEEE International Conference on Image Processing (ICIP)(IEEE 2024) 3730\u20133736.","DOI":"10.1109\/ICIP51287.2024.10648236"},{"key":"e_1_2_12_35_1","first-page":"24604","article-title":"Chip: Channel Independence\u2010Based Pruning for Compact Neural Networks","volume":"34","author":"Sui Y.","year":"2021","journal-title":"Advances in Neural Information Processing Systems"}],"container-title":["IET Image Processing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/ietresearch.onlinelibrary.wiley.com\/doi\/pdf\/10.1049\/ipr2.70231","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/ietresearch.onlinelibrary.wiley.com\/doi\/full-xml\/10.1049\/ipr2.70231","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/ietresearch.onlinelibrary.wiley.com\/doi\/pdf\/10.1049\/ipr2.70231","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,8]],"date-time":"2026-05-08T23:50:08Z","timestamp":1778284208000},"score":1,"resource":{"primary":{"URL":"https:\/\/ietresearch.onlinelibrary.wiley.com\/doi\/10.1049\/ipr2.70231"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,1]]},"references-count":34,"journal-issue":{"issue":"1","published-print":{"date-parts":[[2025,1]]}},"alternative-id":["10.1049\/ipr2.70231"],"URL":"https:\/\/doi.org\/10.1049\/ipr2.70231","archive":["Portico"],"relation":{},"ISSN":["1751-9659","1751-9667"],"issn-type":[{"value":"1751-9659","type":"print"},{"value":"1751-9667","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,1]]},"assertion":[{"value":"2025-05-26","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-10-07","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-10-17","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e70231"}}