{"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/vitae-vision-transformer-advanced-by","title":"ViTAE: Vision Transformer Advanced by Exploring Intrinsic Inductive Bias","arxiv_id":"2106.03348","date":"2021-06-07","proceeding":"NeurIPS 2021 12","authors":["Yufei Xu","Qiming Zhang","Jing Zhang","DaCheng Tao"],"abstract":"Transformers have shown great potential in various computer vision tasks owing to their strong capability in modeling long-range dependency using the self-attention mechanism. Nevertheless, vision transformers treat an image as 1D sequence of visual tokens, lacking an intrinsic inductive bias (IB) in modeling local visual structures and dealing with scale variance. Alternatively, they require large-scale training data and longer training schedules to learn the IB implicitly. In this paper, we propose a novel Vision Transformer Advanced by Exploring intrinsic IB from convolutions, ie, ViTAE. Technically, ViTAE has several spatial pyramid reduction modules to downsample and embed the input image into tokens with rich multi-scale context by using multiple convolutions with different dilation rates. In this way, it acquires an intrinsic scale invariance IB and is able to learn robust feature representation for objects at various scales. Moreover, in each transformer layer, ViTAE has a convolution block in parallel to the multi-head self-attention module, whose features are fused and fed into the feed-forward network. Consequently, it has the intrinsic locality IB and is able to learn local features and global dependencies collaboratively. Experiments on ImageNet as well as downstream tasks prove the superiority of ViTAE over the baseline transformer and concurrent works. Source code and pretrained models will be available at GitHub.","url_abs":"https://arxiv.org/abs/2106.03348v4","url_pdf":"https://arxiv.org/pdf/2106.03348v4.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":"vitae-vision-transformer-advanced-by","repo_url":"https://github.com/Annbless/ViTAE","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"vitae-vision-transformer-advanced-by","repo_url":"https://github.com/ViTAE-Transformer/ViTAE-Transformer","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"inductive-bias","task_name":"Inductive Bias"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"video-object-segmentation","task_name":"Video Object Segmentation"}],"methods":[{"method_slug":"absolute-position-encodings","method_name":"Absolute Position Encodings"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"position-wise-feed-forward-layer","method_name":"Position-Wise Feed-Forward Layer"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"transformer","method_name":"Transformer"},{"method_slug":"vision-transformer","method_name":"Vision Transformer"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViTAE-B-Stage","rank_in_archive_order":414,"of":1060,"metrics":{"GFLOPs":"27.6","Number of params":"48.5M","Top 1 Accuracy":"83.6%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViTAE-S-Stage","rank_in_archive_order":561,"of":1060,"metrics":{"GFLOPs":"12.0","Number of params":"19.2M","Top 1 Accuracy":"82.2%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViTAE-13M","rank_in_archive_order":670,"of":1060,"metrics":{"GFLOPs":"6.8","Number of params":"13.2M","Top 1 Accuracy":"81%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViTAE-6M","rank_in_archive_order":860,"of":1060,"metrics":{"GFLOPs":"4","Number of params":"6.5M","Top 1 Accuracy":"77.9%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViTAE-T-Stage","rank_in_archive_order":897,"of":1060,"metrics":{"GFLOPs":"4.6","Number of params":"4.8M","Top 1 Accuracy":"76.8%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ViTAE-T","rank_in_archive_order":952,"of":1060,"metrics":{"GFLOPs":"3.0","Top 1 Accuracy":"75.3%"},"uses_additional_data":false},{"leaderboard":"/sota/video-object-segmentation-on-davis-2016","task":"Video Object Segmentation","dataset":"DAVIS 2016","model":"ViTAE-T-Stage","rank_in_archive_order":13,"of":24,"metrics":{"F-Score":"90.4","J&F":"89.8","Jaccard (Mean)":"89.2"},"uses_additional_data":false},{"leaderboard":"/sota/video-object-segmentation-on-davis-2017","task":"Video Object Segmentation","dataset":"DAVIS 2017","model":"ViTAE-T-Stage","rank_in_archive_order":2,"of":5,"metrics":{"F-Score":"85.5","J&F":"82.5","Jaccard (Mean)":"79.4"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2106.03348","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2106.03348"}},"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/Annbless/ViTAE","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ViTAE-Transformer/ViTAE-Transformer","reach":{"status":"ok"}}],"summary":{"ran_fixture":1,"ran_draft_wrong":1},"by_repo_kind":{},"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":2,"samples":[{"code_sha256_prefix":"61c027be5178a479","entry":"window_partition","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_fixture","verification_level":2,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"61c027be5178a479"}},{"code_sha256_prefix":"5e2dbd713ec13547","entry":"window_reverse","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"ran_draft_wrong","verification_level":2,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"5e2dbd713ec13547"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}