{"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/x-volution-on-the-unification-of-convolution","title":"X-volution: On the unification of convolution and self-attention","arxiv_id":"2106.02253","date":"2021-06-04","proceeding":null,"authors":["Xuanhong Chen","Hang Wang","Bingbing Ni"],"abstract":"Convolution and self-attention are acting as two fundamental building blocks in deep neural networks, where the former extracts local image features in a linear way while the latter non-locally encodes high-order contextual relationships. Though essentially complementary to each other, i.e., first-/high-order, stat-of-the-art architectures, i.e., CNNs or transformers lack a principled way to simultaneously apply both operations in a single computational module, due to their heterogeneous computing pattern and excessive burden of global dot-product for visual tasks. In this work, we theoretically derive a global self-attention approximation scheme, which approximates a self-attention via the convolution operation on transformed features. Based on the approximated scheme, we establish a multi-branch elementary module composed of both convolution and self-attention operation, capable of unifying both local and non-local feature interaction. Importantly, once trained, this multi-branch module could be conditionally converted into a single standard convolution operation via structural re-parameterization, rendering a pure convolution styled operator named X-volution, ready to be plugged into any modern networks as an atomic operation. Extensive experiments demonstrate that the proposed X-volution, achieves highly competitive visual understanding improvements (+1.2% top-1 accuracy on ImageNet classification, +1.7 box AP and +1.5 mask AP on COCO detection and segmentation).","url_abs":"https://arxiv.org/abs/2106.02253v2","url_pdf":"https://arxiv.org/pdf/2106.02253v2.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":[],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"instance-segmentation","task_name":"Instance Segmentation"},{"task_slug":"object-detection","task_name":"Object Detection"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ResNet-50 (X-volution, stage3)","rank_in_archive_order":908,"of":1060,"metrics":{"Top 1 Accuracy":"76.6%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"ResNet-34 (X-volution, stage3)","rank_in_archive_order":960,"of":1060,"metrics":{"Top 1 Accuracy":"75%"},"uses_additional_data":false},{"leaderboard":"/sota/instance-segmentation-on-coco-minival","task":"Instance Segmentation","dataset":"COCO minival","model":"Mask R-CNN (FPN, X-volution, SA)","rank_in_archive_order":85,"of":93,"metrics":{"APL":"53.1","APM":"40","APS":"19.2","mask AP":"37.2"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-coco-minival","task":"Object Detection","dataset":"COCO minival","model":"Faster R-CNN (FPN, X-volution)","rank_in_archive_order":149,"of":220,"metrics":{"AP50":"64","AP75":"46.4","APL":"55","APM":"46","APS":"26.9","box AP":"42.8"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2106.02253","atlas_url":"https://app.syntology.ai/?focus=2106.02253","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}