{"about":{"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.","site":"https://codewithpapers.app","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","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/paper/uniformer-unified-transformer-for-efficient","title":"UniFormer: Unified Transformer for Efficient Spatial-Temporal Representation Learning","arxiv_id":null,"date":"2021-09-29","proceeding":"ICLR 2022 4","authors":["Kunchang Li","Yali Wang","Gao Peng","Guanglu Song","Yu Liu","Hongsheng Li","Yu Qiao"],"abstract":"It is a challenging task to learn rich and multi-scale spatial-temporal semantics from high-dimensional videos, due to large local redundancy and complex global dependency between video frames. The recent advances in this research have been mainly driven by 3D convolutional neural networks and vision transformers. Although 3D convolution can efficiently aggregate local context to suppress local redundancy from a small 3D neighborhood, it lacks the capability to capture global dependency because of the limited receptive field. Alternatively, vision transformers can effectively capture long-range dependency by self-attention mechanism, while having limitations on reducing local redundancy with blind similarity comparison among all the tokens in each layer. Based on these observations, we propose a novel Unified transFormer (UniFormer) which seamlessly integrates merits of 3D convolution and spatial-temporal self-attention in a concise transformer format, and achieves a preferable balance between computation and accuracy. Different from traditional transformers, our relation aggregator can tackle both spatial-temporal redundancy and dependency, by learning local and global token affinity respectively in shallow and deep layers. We conduct extensive experiments on the popular video benchmarks, e.g., Kinetics-400, Kinetics-600, and Something-Something V1&V2. With only ImageNet-1K pretraining, our UniFormer achieves 82.9%/84.8% top-1 accuracy on Kinetics-400/Kinetics-600, while requiring 10x fewer GFLOPs than other state-of-the-art methods. For Something-Something V1 and V2, our UniFormer achieves new state-of-the-art performances of 60.8% and 71.4% top-1 accuracy respectively. ","url_abs":"https://openreview.net/forum?id=nBU_u6DLvoK","url_pdf":"https://openreview.net/pdf?id=nBU_u6DLvoK","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":"uniformer-unified-transformer-for-efficient","repo_url":"https://github.com/sense-x/uniformer","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"uniformer-unified-transformer-for-efficient","repo_url":"https://github.com/lucidrains/uniformer-pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null},{"paper_slug":"uniformer-unified-transformer-for-efficient","repo_url":"https://github.com/towhee-io/towhee","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":"representation-learning","task_name":"Representation Learning"}],"methods":[{"method_slug":"3d-convolution","method_name":"3D Convolution"},{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/action-classification-on-kinetics-400","task":"Action Classification","dataset":"Kinetics-400","model":"UniFormer-B (ImageNet-1K)","rank_in_archive_order":71,"of":207,"metrics":{"Acc@1":"82.9","Acc@5":"94.5","FLOPs (G) x views":"259x4"},"uses_additional_data":false},{"leaderboard":"/sota/action-classification-on-kinetics-600","task":"Action Classification","dataset":"Kinetics-600","model":"UniFormer-B (ImageNet-1K)","rank_in_archive_order":31,"of":65,"metrics":{"GFLOPs":"259x4","Top-1 Accuracy":"84.8","Top-5 Accuracy":"96.7"},"uses_additional_data":true},{"leaderboard":"/sota/action-recognition-in-videos-on-something-1","task":"Action Recognition","dataset":"Something-Something V1","model":"UniFormer-B (IN-1K + Kinetics400)","rank_in_archive_order":9,"of":74,"metrics":{"GFLOPs":"259x3","Param.":"50.1","Top 1 Accuracy":"60.9","Top 5 Accuracy":"87.3"},"uses_additional_data":false},{"leaderboard":"/sota/action-recognition-in-videos-on-something-1","task":"Action Recognition","dataset":"Something-Something V1","model":"UniFormer-B (IN-1K + Kinetics600)","rank_in_archive_order":13,"of":74,"metrics":{"GFLOPs":"41.8x3","Param.":"21.4","Top 1 Accuracy":"57.6","Top 5 Accuracy":"84.9"},"uses_additional_data":false},{"leaderboard":"/sota/action-recognition-in-videos-on-something","task":"Action Recognition","dataset":"Something-Something V2","model":"UniFormer-B (IN-1K + Kinetics400 pretrain)","rank_in_archive_order":31,"of":123,"metrics":{"GFLOPs":"259x3","Parameters":"50.1","Top-1 Accuracy":"71.2","Top-5 Accuracy":"92.8"},"uses_additional_data":true},{"leaderboard":"/sota/action-recognition-in-videos-on-something","task":"Action Recognition","dataset":"Something-Something V2","model":"UniFormer-S (IN-1K + Kinetics600 pretrain)","rank_in_archive_order":47,"of":123,"metrics":{"GFLOPs":"41.8x3","Parameters":"21.4","Top-1 Accuracy":"69.4","Top-5 Accuracy":"92.1"},"uses_additional_data":true}],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers","held":null,"reason":"no_arxiv_id","checked_against":null,"by_repo":[{"repo":"sense-x/uniformer","kind":"official","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":1,"harvested_for":["2201.04676"],"harvested_for_calls":[{"arxiv_id":"2201.04676","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2201.04676"},"call":"get_harvested_code_for_paper(arxiv_id=\"2201.04676\")"}]},{"repo":"lucidrains/uniformer-pytorch","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":1,"harvested_for":["2201.04676"],"harvested_for_calls":[{"arxiv_id":"2201.04676","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2201.04676"},"call":"get_harvested_code_for_paper(arxiv_id=\"2201.04676\")"}]},{"repo":"towhee-io/towhee","kind":"listed","samples":0,"ran":0,"instrument":0,"unverified":0,"harvested_for_other_papers":12,"harvested_for":["1512.03385","2103.03206","2103.08784","2103.14030","2104.11227","2201.03545","2201.04676","2201.04850","2201.08377","2203.16518","2204.01697","2206.10436"],"harvested_for_calls":[{"arxiv_id":"1512.03385","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"1512.03385"},"call":"get_harvested_code_for_paper(arxiv_id=\"1512.03385\")"},{"arxiv_id":"2103.03206","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2103.03206"},"call":"get_harvested_code_for_paper(arxiv_id=\"2103.03206\")"},{"arxiv_id":"2103.08784","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2103.08784"},"call":"get_harvested_code_for_paper(arxiv_id=\"2103.08784\")"},{"arxiv_id":"2103.14030","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2103.14030"},"call":"get_harvested_code_for_paper(arxiv_id=\"2103.14030\")"},{"arxiv_id":"2104.11227","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2104.11227"},"call":"get_harvested_code_for_paper(arxiv_id=\"2104.11227\")"},{"arxiv_id":"2201.03545","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2201.03545"},"call":"get_harvested_code_for_paper(arxiv_id=\"2201.03545\")"},{"arxiv_id":"2201.04676","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2201.04676"},"call":"get_harvested_code_for_paper(arxiv_id=\"2201.04676\")"},{"arxiv_id":"2201.04850","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2201.04850"},"call":"get_harvested_code_for_paper(arxiv_id=\"2201.04850\")"},{"arxiv_id":"2201.08377","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2201.08377"},"call":"get_harvested_code_for_paper(arxiv_id=\"2201.08377\")"},{"arxiv_id":"2203.16518","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2203.16518"},"call":"get_harvested_code_for_paper(arxiv_id=\"2203.16518\")"},{"arxiv_id":"2204.01697","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2204.01697"},"call":"get_harvested_code_for_paper(arxiv_id=\"2204.01697\")"},{"arxiv_id":"2206.10436","tool":"get_harvested_code_for_paper","arguments":{"arxiv_id":"2206.10436"},"call":"get_harvested_code_for_paper(arxiv_id=\"2206.10436\")"}]}],"by_repo_harvested_for_is":"harvested_for: the arXiv ids (as Syntology's graph holds them) of every paper whose samples Syntology's graph links to this repository, in arXiv-id order, complete; the page names the first 3 and says how many more. A link only: it does not say the papers are the same work, that one copied the other, or that the code implements this paper's method; harvested_for_calls gives the get_harvested_code_for_paper call for each arXiv-shaped id among them","by_repo_is":"one row per repository in the archive's code links, Syntology's graph links or the samples; samples 0 means no sample is linked to this paper, a statement about Syntology's coverage, not about the repository: with harvested_for_other_papers present, Syntology harvested the repository for that many other papers, and without it, Syntology harvested nothing from it; placed_by_identical_code counts samples whose link records no repository, placed here because identical code was harvested from this repository for another paper; ran counts executions on a synthesized input and instrument counts failures of Syntology's instrument, not of the code"},"code_links_note":null,"arxiv_metadata":null,"syntology_extracted_results":null,"record_sha256":"f5116acbace138b2b4630c354ff9dca5b3ab116c314af47786818ca91278e6c6","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}