{"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/omni-sourced-webly-supervised-learning-for","title":"Omni-sourced Webly-supervised Learning for Video Recognition","arxiv_id":"2003.13042","date":"2020-03-29","proceeding":"ECCV 2020 8","authors":["Haodong Duan","Yue Zhao","Yuanjun Xiong","Wentao Liu","Dahua Lin"],"abstract":"We introduce OmniSource, a novel framework for leveraging web data to train video recognition models. OmniSource overcomes the barriers between data formats, such as images, short videos, and long untrimmed videos for webly-supervised learning. First, data samples with multiple formats, curated by task-specific data collection and automatically filtered by a teacher model, are transformed into a unified form. Then a joint-training strategy is proposed to deal with the domain gaps between multiple data sources and formats in webly-supervised learning. Several good practices, including data balancing, resampling, and cross-dataset mixup are adopted in joint training. Experiments show that by utilizing data from multiple sources and formats, OmniSource is more data-efficient in training. With only 3.5M images and 800K minutes videos crawled from the internet without human labeling (less than 2% of prior works), our models learned with OmniSource improve Top-1 accuracy of 2D- and 3D-ConvNet baseline models by 3.0% and 3.9%, respectively, on the Kinetics-400 benchmark. With OmniSource, we establish new records with different pretraining strategies for video recognition. Our best models achieve 80.4%, 80.5%, and 83.6 Top-1 accuracies on the Kinetics-400 benchmark respectively for training-from-scratch, ImageNet pre-training and IG-65M pre-training.","url_abs":"https://arxiv.org/abs/2003.13042v2","url_pdf":"https://arxiv.org/pdf/2003.13042v2.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":"omni-sourced-webly-supervised-learning-for","repo_url":"https://github.com/open-mmlab/mmaction","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"omni-sourced-webly-supervised-learning-for","repo_url":"https://github.com/MCG-NJU/CPD-Video","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"omni-sourced-webly-supervised-learning-for","repo_url":"https://github.com/open-mmlab/mmaction2","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"action-classification","task_name":"Action Classification"},{"task_slug":"action-recognition-in-videos","task_name":"Action Recognition"},{"task_slug":"video-recognition","task_name":"Video Recognition"}],"methods":[{"method_slug":"mixup","method_name":"Mixup"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/action-classification-on-kinetics-400","task":"Action Classification","dataset":"Kinetics-400","model":"OmniSource irCSN-152 (IG-Kinetics-65M pretrain)","rank_in_archive_order":66,"of":207,"metrics":{"Acc@1":"83.6"},"uses_additional_data":false},{"leaderboard":"/sota/action-classification-on-kinetics-400","task":"Action Classification","dataset":"Kinetics-400","model":"OmniSource SlowOnly R101 8x8(ImageNet pretrain)","rank_in_archive_order":98,"of":207,"metrics":{"Acc@1":"80.5","Acc@5":"94.4"},"uses_additional_data":false},{"leaderboard":"/sota/action-classification-on-kinetics-400","task":"Action Classification","dataset":"Kinetics-400","model":"OmniSource SlowOnly R101 8x8 (Scratch)","rank_in_archive_order":102,"of":207,"metrics":{"Acc@1":"80.4","Acc@5":"94.4"},"uses_additional_data":false},{"leaderboard":"/sota/action-recognition-in-videos-on-hmdb-51","task":"Action Recognition","dataset":"HMDB-51","model":"OmniSource (SlowOnly-8x8-R101-RGB + I3D Flow)","rank_in_archive_order":10,"of":77,"metrics":{"Average accuracy of 3 splits":"83.8"},"uses_additional_data":true},{"leaderboard":"/sota/action-recognition-in-videos-on-ucf101","task":"Action Recognition","dataset":"UCF101","model":"OmniSource (SlowOnly-8x8-R101-RGB + I3D-Flow)","rank_in_archive_order":7,"of":91,"metrics":{"3-fold Accuracy":"98.6"},"uses_additional_data":true}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2003.13042","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2003.13042"}},"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/open-mmlab/mmaction","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MCG-NJU/CPD-Video","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/open-mmlab/mmaction2","reach":null}],"summary":{"unverified":13},"by_repo_kind":{"official":{"samples":13,"ran":0,"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":0,"samples":[{"code_sha256_prefix":"4bed2534f3a2204f","entry":"SSIM_loss","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/losses/flow_losses.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/losses/flow_losses.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"4bed2534f3a2204f"}},{"code_sha256_prefix":"8499456603e86421","entry":"anchor_target_single","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/core/anchor2d/anchor_target.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/core/anchor2d/anchor_target.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"8499456603e86421"}},{"code_sha256_prefix":"d65ea05c493bc2d8","entry":"bbox_overlaps","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/core/bbox2d/geometry.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/core/bbox2d/geometry.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"d65ea05c493bc2d8"}},{"code_sha256_prefix":"bea6a74545a6b1b9","entry":"bbox_target_single","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/core/bbox2d/bbox_target.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/core/bbox2d/bbox_target.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"bea6a74545a6b1b9"}},{"code_sha256_prefix":"d1055ad5929a8347","entry":"charbonnier_loss","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/losses/flow_losses.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/losses/flow_losses.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"d1055ad5929a8347"}},{"code_sha256_prefix":"d05f21f2da402e1c","entry":"classwise_regression_loss","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/losses/ssn_losses.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/losses/ssn_losses.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"d05f21f2da402e1c"}},{"code_sha256_prefix":"66a6fb33d71062f2","entry":"completeness_loss","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/losses/ssn_losses.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/losses/ssn_losses.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"66a6fb33d71062f2"}},{"code_sha256_prefix":"a57ba57f6dde621b","entry":"expand_target","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/core/bbox2d/bbox_target.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/core/bbox2d/bbox_target.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"a57ba57f6dde621b"}},{"code_sha256_prefix":"8fa249eba3a0ce57","entry":"images_to_levels","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/core/anchor2d/anchor_target.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/core/anchor2d/anchor_target.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"8fa249eba3a0ce57"}},{"code_sha256_prefix":"6fb204c4574e77ac","entry":"temporal_iou","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/core/bbox1d/geometry.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/core/bbox1d/geometry.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"6fb204c4574e77ac"}},{"code_sha256_prefix":"380e288d5962b717","entry":"weighted_binary_cross_entropy","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/losses/losses.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/losses/losses.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"380e288d5962b717"}},{"code_sha256_prefix":"8a052a08051eb8bb","entry":"weighted_cross_entropy","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/losses/losses.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/losses/losses.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"8a052a08051eb8bb"}},{"code_sha256_prefix":"48433fd136b26537","entry":"weighted_nll_loss","repo":"open-mmlab/mmaction","repo_kind":"official","path":"mmaction/losses/losses.py","file_url":"https://github.com/open-mmlab/mmaction/blob/HEAD/mmaction/losses/losses.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"48433fd136b26537"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}