{"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/separable-self-attention-for-mobile-vision","title":"Separable Self-attention for Mobile Vision Transformers","arxiv_id":"2206.02680","date":"2022-06-06","proceeding":null,"authors":["Sachin Mehta","Mohammad Rastegari"],"abstract":"Mobile vision transformers (MobileViT) can achieve state-of-the-art performance across several mobile vision tasks, including classification and detection. Though these models have fewer parameters, they have high latency as compared to convolutional neural network-based models. The main efficiency bottleneck in MobileViT is the multi-headed self-attention (MHA) in transformers, which requires $O(k^2)$ time complexity with respect to the number of tokens (or patches) $k$. Moreover, MHA requires costly operations (e.g., batch-wise matrix multiplication) for computing self-attention, impacting latency on resource-constrained devices. This paper introduces a separable self-attention method with linear complexity, i.e. $O(k)$. A simple yet effective characteristic of the proposed method is that it uses element-wise operations for computing self-attention, making it a good choice for resource-constrained devices. The improved model, MobileViTv2, is state-of-the-art on several mobile vision tasks, including ImageNet object classification and MS-COCO object detection. With about three million parameters, MobileViTv2 achieves a top-1 accuracy of 75.6% on the ImageNet dataset, outperforming MobileViT by about 1% while running $3.2\\times$ faster on a mobile device. Our source code is available at: \\url{https://github.com/apple/ml-cvnets}","url_abs":"https://arxiv.org/abs/2206.02680v1","url_pdf":"https://arxiv.org/pdf/2206.02680v1.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":"separable-self-attention-for-mobile-vision","repo_url":"https://github.com/apple/ml-cvnets","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"separable-self-attention-for-mobile-vision","repo_url":"https://github.com/IMvision12/keras-vision-models","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"separable-self-attention-for-mobile-vision","repo_url":"https://github.com/jaiwei98/mobile-vit-pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"separable-self-attention-for-mobile-vision","repo_url":"https://github.com/rwightman/pytorch-image-models","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"separable-self-attention-for-mobile-vision","repo_url":"https://github.com/t0nyliang/EEGMobile","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"separable-self-attention-for-mobile-vision","repo_url":"https://github.com/leondgarse/keras_cv_attention_models/tree/main/keras_cv_attention_models/mobilevit","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":null},{"paper_slug":"separable-self-attention-for-mobile-vision","repo_url":"https://github.com/mindspore-courses/External-Attention-MindSpore/blob/main/model/attention/MobileViTv2Attention.py","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"separable-self-attention-for-mobile-vision","repo_url":"https://gitlab.com/birder/birder","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"object-detection","task_name":"Object Detection"}],"methods":[{"method_slug":"mobilevit","method_name":"MobileViT"},{"method_slug":"mobilevitv2","method_name":"MobileViTv2"}],"datasets_introduced":[],"methods_introduced":[{"slug":"mobilevitv2","name":"MobileViTv2","full_name":"MobileViTv2"}],"results":[{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"MobileViTv2-1.0","rank_in_archive_order":852,"of":1060,"metrics":{"GFLOPs":"1.8","Number of params":"4.9M","Top 1 Accuracy":"78.1%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"MobileViTv2-0.75","rank_in_archive_order":943,"of":1060,"metrics":{"GFLOPs":"1.0","Number of params":"2.9M","Top 1 Accuracy":"75.6%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"MobileViTv2-0.5","rank_in_archive_order":1023,"of":1060,"metrics":{"GFLOPs":"0.5","Number of params":"1.4M","Top 1 Accuracy":"70.2%"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2206.02680","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2206.02680"}},"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/rwightman/pytorch-image-models","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/apple/ml-cvnets","reach":{"status":"ok","spdx":"NOASSERTION"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/leondgarse/keras_cv_attention_models/tree/main/keras_cv_attention_models/mobilevit","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/IMvision12/keras-vision-models","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://gitlab.com/birder/birder","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/mindspore-courses/External-Attention-MindSpore/blob/main/model/attention/MobileViTv2Attention.py","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/jaiwei98/mobile-vit-pytorch","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/t0nyliang/EEGMobile","reach":{"status":"ok"}}],"summary":{"ran_draft_wrong":1},"by_repo_kind":{"listed":{"samples":1,"ran":1,"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":1,"samples":[{"code_sha256_prefix":"46467acb70052484","entry":"conv_2d","repo":"jaiwei98/mobile-vit-pytorch","repo_kind":"listed","path":"mobile_vit/mobilevit.py","file_url":"https://github.com/jaiwei98/mobile-vit-pytorch/blob/HEAD/mobile_vit/mobilevit.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"46467acb70052484"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}