{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/variable-rate-image-compression-with","title":"Variable Rate Image Compression with Recurrent Neural Networks","arxiv_id":"1511.06085","date":"2015-11-19","proceeding":null,"authors":["George Toderici","Sean M. O'Malley","Sung Jin Hwang","Damien Vincent","David Minnen","Shumeet Baluja","Michele Covell","Rahul Sukthankar"],"abstract":"A large fraction of Internet traffic is now driven by requests from mobile\ndevices with relatively small screens and often stringent bandwidth\nrequirements. Due to these factors, it has become the norm for modern\ngraphics-heavy websites to transmit low-resolution, low-bytecount image\npreviews (thumbnails) as part of the initial page load process to improve\napparent page responsiveness. Increasing thumbnail compression beyond the\ncapabilities of existing codecs is therefore a current research focus, as any\nbyte savings will significantly enhance the experience of mobile device users.\nToward this end, we propose a general framework for variable-rate image\ncompression and a novel architecture based on convolutional and deconvolutional\nLSTM recurrent networks. Our models address the main issues that have prevented\nautoencoder neural networks from competing with existing image compression\nalgorithms: (1) our networks only need to be trained once (not per-image),\nregardless of input image dimensions and the desired compression rate; (2) our\nnetworks are progressive, meaning that the more bits are sent, the more\naccurate the image reconstruction; and (3) the proposed architecture is at\nleast as efficient as a standard purpose-trained autoencoder for a given number\nof bits. On a large-scale benchmark of 32$\\times$32 thumbnails, our LSTM-based\napproaches provide better visual quality than (headerless) JPEG, JPEG2000 and\nWebP, with a storage size that is reduced by 10% or more.","url_abs":"http://arxiv.org/abs/1511.06085v5","url_pdf":"http://arxiv.org/pdf/1511.06085v5.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"variable-rate-image-compression-with","repo_url":"https://github.com/alexandru-dinu/cae","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"variable-rate-image-compression-with","repo_url":"https://github.com/eyalfish/artoveq","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"image-compression","task_name":"Image Compression"},{"task_slug":"image-reconstruction","task_name":"Image Reconstruction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1511.06085","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1511.06085"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/eyalfish/artoveq","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/alexandru-dinu/cae","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_violates":1,"ran_draft_wrong":1,"unverified":1},"by_repo_kind":{"listed":{"samples":3,"ran":2,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":3,"samples":[{"code_sha256_prefix":"8e31b020598d799c","entry":"get_accuracy","repo":"eyalfish/artoveq","repo_kind":"listed","path":"ResidualVQVAE.py","file_url":"https://github.com/eyalfish/artoveq/blob/HEAD/ResidualVQVAE.py","link_basis":"first_harvest_node","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"8e31b020598d799c"}},{"code_sha256_prefix":"86301a36fc05d312","entry":"init_weights_lbg","repo":"eyalfish/artoveq","repo_kind":"listed","path":"adaptivecb.py","file_url":"https://github.com/eyalfish/artoveq/blob/HEAD/adaptivecb.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"86301a36fc05d312"}},{"code_sha256_prefix":"8101a561f67542dc","entry":"train_classifier","repo":"eyalfish/artoveq","repo_kind":"listed","path":"ResidualVQVAE.py","file_url":"https://github.com/eyalfish/artoveq/blob/HEAD/ResidualVQVAE.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"8101a561f67542dc"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}