{"url":"/method/coordinate-attention","slug":"coordinate-attention","name":"Coordinate attention","full_name":"Coordinate attention","full_name_withheld":false,"description_markdown":"Hou et al. proposed coordinate attention,\r\na novel attention mechanism which\r\nembeds positional information into channel attention,\r\nso that the network can focus on large important regions \r\nat little computational cost.\r\n\r\nThe coordinate attention mechanism has two consecutive steps, coordinate information embedding and coordinate attention generation. First, two spatial extents of pooling kernels encode each channel horizontally  and  vertically. In the second step, a shared $1\\times 1$ convolutional transformation function is applied to the concatenated outputs of the two pooling layers. Then coordinate attention splits the resulting tensor into two separate tensors to yield attention vectors with the same number of channels for horizontal and vertical coordinates of the  input $X$ along. This can be written as \r\n\\begin{align}\r\n    z^h &= \\text{GAP}^h(X) \r\n\\end{align}\r\n\\begin{align}\r\n    z^w &= \\text{GAP}^w(X)\r\n\\end{align}\r\n\\begin{align}\r\n    f &= \\delta(\\text{BN}(\\text{Conv}_1^{1\\times 1}([z^h;z^w])))\r\n\\end{align}\r\n\\begin{align}\r\n    f^h, f^w &= \\text{Split}(f)\r\n\\end{align}\r\n\\begin{align}\r\n    s^h &= \\sigma(\\text{Conv}_h^{1\\times 1}(f^h))\r\n\\end{align}\r\n\\begin{align}\r\n    s^w &= \\sigma(\\text{Conv}_w^{1\\times 1}(f^w))\r\n\\end{align}\r\n\\begin{align}\r\n    Y &= X s^h  s^w\r\n\\end{align}\r\nwhere $\\text{GAP}^h$ and $\\text{GAP}^w$ denote pooling functions for vertical and horizontal coordinates, and $s^h \\in \\mathbb{R}^{C\\times 1\\times W}$ and $s^w \\in \\mathbb{R}^{C\\times H\\times 1}$ represent corresponding attention weights. \r\n\r\nUsing coordinate attention, the network can accurately obtain the position of a targeted object.\r\nThis approach has a larger receptive field than BAM and CBAM.\r\nLike an SE block, it also models cross-channel relationships, effectively enhancing the expressive power of the learned features.\r\nDue to its lightweight design and flexibility, \r\nit can be easily used in classical building blocks of mobile networks.","description_state":"present","introduced_year":null,"introduced_by":{"title":"Coordinate Attention for Efficient Mobile Network Design","paper":"/paper/coordinate-attention-for-efficient-mobile","first_author":"Qibin Hou","n_authors":3,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/coordinate-attention-for-efficient-mobile"},"source":{"url":"https://arxiv.org/abs/2103.02907v1","title":"Coordinate Attention for Efficient Mobile Network Design","url_on_a_paper_host":true},"code_snippet_url":null,"code_snippet_url_on_a_code_host":false,"categories":[{"area":"General","area_id":"general","collection":"Attention Mechanisms","url":"/methods/category/attention-mechanisms","pwc_aliases":["attention-mechanisms-1"]}],"n_papers_tagged":31,"archive_num_papers":31,"papers_newest_first":[{"paper":null,"title":"Design description of Wisdom Computing Persperctive","date":"2025-05-02","arxiv_id":"2505.03800","n_code_links":0,"syntology":null},{"paper":null,"title":"Enhancing Traffic Sign Recognition On The Performance Based On Yolov8","date":"2025-04-02","arxiv_id":"2504.02884","n_code_links":0,"syntology":null},{"paper":"/paper/lightendostereo-a-real-time-lightweight","title":"LightEndoStereo: A Real-time Lightweight Stereo Matching Method for Endoscopy Images","date":"2025-03-02","arxiv_id":"2503.00731","n_code_links":1,"syntology":null},{"paper":"/paper/ptychoformer-a-physics-guided-deep-learning","title":"A Physics-Inspired Deep Learning Framework with Polar Coordinate Attention for Ptychographic Imaging","date":"2024-11-25","arxiv_id":"2412.06806","n_code_links":1,"syntology":null},{"paper":"/paper/hyperspectral-imaging-based-perception-in","title":"Hyperspectral Imaging-Based Perception in Autonomous Driving Scenarios: Benchmarking Baseline Semantic Segmentation Models","date":"2024-10-29","arxiv_id":"2410.22101","n_code_links":1,"syntology":null},{"paper":null,"title":"Optimizing YOLO Architectures for Optimal Road Damage Detection and Classification: A Comparative Study from YOLOv7 to YOLOv10","date":"2024-10-10","arxiv_id":"2410.08409","n_code_links":0,"syntology":null},{"paper":null,"title":"Improved Unet model for brain tumor image segmentation based on ASPP-coordinate attention mechanism","date":"2024-09-13","arxiv_id":"2409.08588","n_code_links":0,"syntology":null},{"paper":null,"title":"RICAU-Net: Residual-block Inspired Coordinate Attention U-Net for Segmentation of Small and Sparse Calcium Lesions in Cardiac CT","date":"2024-09-11","arxiv_id":"2409.06993","n_code_links":0,"syntology":null},{"paper":"/paper/alss-yolo-an-adaptive-lightweight-channel","title":"ALSS-YOLO: An Adaptive Lightweight Channel Split and Shuffling Network for TIR Wildlife Detection in UAV Imagery","date":"2024-09-10","arxiv_id":"2409.06259","n_code_links":1,"syntology":null},{"paper":null,"title":"CSANet: Channel Spatial Attention Network for Robust 3D Face Alignment and Reconstruction","date":"2024-05-30","arxiv_id":"2405.19659","n_code_links":0,"syntology":null},{"paper":null,"title":"ELA: Efficient Local Attention for Deep Convolutional Neural Networks","date":"2024-03-02","arxiv_id":"2403.01123","n_code_links":0,"syntology":null},{"paper":"/paper/deep-linear-array-pushbroom-image-restoration","title":"Deep Linear Array Pushbroom Image Restoration: A Degradation Pipeline and Jitter-Aware Restoration Network","date":"2024-01-16","arxiv_id":"2401.08171","n_code_links":1,"syntology":{"ran":5,"of":8,"unverified":3,"pointer_only":0}},{"paper":null,"title":"YOLO algorithm with hybrid attention feature pyramid network for solder joint defect detection","date":"2024-01-02","arxiv_id":"2401.01214","n_code_links":0,"syntology":null},{"paper":null,"title":"YOLOv5s-BC: An improved YOLOv5s-based method for real-time apple detection","date":"2023-11-10","arxiv_id":"2311.05811","n_code_links":0,"syntology":null},{"paper":null,"title":"Marine Debris Detection in Satellite Surveillance using Attention Mechanisms","date":"2023-07-09","arxiv_id":"2307.04128","n_code_links":0,"syntology":null},{"paper":null,"title":"Multi-cropping Contrastive Learning and Domain Consistency for Unsupervised Image-to-Image Translation","date":"2023-04-24","arxiv_id":"2304.12235","n_code_links":0,"syntology":null},{"paper":null,"title":"Two-stage MR Image Segmentation Method for Brain Tumors based on Attention Mechanism","date":"2023-04-17","arxiv_id":"2304.08072","n_code_links":0,"syntology":null},{"paper":null,"title":"Fast vehicle detection algorithm based on lightweight YOLO7-tiny","date":"2023-04-12","arxiv_id":"2304.06002","n_code_links":0,"syntology":null},{"paper":null,"title":"PCCA-Model: an attention module for medical image segmentation","date":"2023-04-01","arxiv_id":null,"n_code_links":0,"syntology":null},{"paper":null,"title":"TWR-MCAE: A Data Augmentation Method for Through-the-Wall Radar Human Motion Recognition","date":"2023-01-06","arxiv_id":"2301.02488","n_code_links":0,"syntology":null},{"paper":null,"title":"Road Damages Detection and Classification with YOLOv7","date":"2022-10-31","arxiv_id":"2211.00091","n_code_links":0,"syntology":null},{"paper":"/paper/aia-attention-in-attention-within-collaborate","title":"AIA: Attention in Attention Within Collaborate Domains","date":"2022-10-07","arxiv_id":null,"n_code_links":1,"syntology":null},{"paper":null,"title":"A New Method on Mask-Wearing Detection for Natural Population Based on Improved YOLOv4","date":"2022-08-24","arxiv_id":"2208.11353","n_code_links":0,"syntology":null},{"paper":null,"title":"Improved lightweight identification of agricultural diseases based on MobileNetV3","date":"2022-07-19","arxiv_id":"2207.11238","n_code_links":0,"syntology":null},{"paper":"/paper/an-improved-lightweight-yolov5-model-based-on","title":"An Improved Lightweight YOLOv5 Model Based on Attention Mechanism for Face Mask Detection","date":"2022-03-30","arxiv_id":"2203.16506","n_code_links":1,"syntology":null},{"paper":null,"title":"SAR-ShipNet: SAR-Ship Detection Neural Network via Bidirectional Coordinate Attention and Multi-resolution Feature Fusion","date":"2022-03-29","arxiv_id":"2203.15480","n_code_links":0,"syntology":null},{"paper":null,"title":"Improving Generalization of Deep Networks for Estimating Physical Properties of Containers and Fillings","date":"2022-03-02","arxiv_id":"2203.01192","n_code_links":0,"syntology":null},{"paper":null,"title":"A Lightweight Dual-Domain Attention Framework for Sparse-View CT Reconstruction","date":"2022-02-19","arxiv_id":"2202.09609","n_code_links":0,"syntology":null},{"paper":null,"title":"Attention-Guided NIR Image Colorization via Adaptive Fusion of Semantic and Texture Clues","date":"2021-07-20","arxiv_id":"2107.09237","n_code_links":0,"syntology":null},{"paper":null,"title":"Joint Attention for Multi-Agent Coordination and Social Learning","date":"2021-04-15","arxiv_id":"2104.07750","n_code_links":0,"syntology":null}],"papers_shown":30,"tasks":[{"task":"/task/semantic-segmentation","name":"Semantic Segmentation","papers":6},{"task":"/task/data-augmentation","name":"Data Augmentation","papers":4},{"task":"/task/object-detection","name":"Object Detection","papers":4},{"task":"/task/image-segmentation","name":"Image Segmentation","papers":3},{"task":"/task/segmentation","name":"Segmentation","papers":3},{"task":"/task/object-detection-1","name":"object-detection","papers":3},{"task":"/task/autonomous-driving","name":"Autonomous Driving","papers":2},{"task":"/task/deep-learning","name":"Deep Learning","papers":2},{"task":null,"name":"Generative Adversarial Network","papers":2},{"task":"/task/mamba","name":"Mamba","papers":2},{"task":"/task/road-damage-detection","name":"Road Damage Detection","papers":2},{"task":"/task/3d-face-alignment","name":"3D Face Alignment","papers":1},{"task":"/task/autonomous-vehicles","name":"Autonomous Vehicles","papers":1},{"task":"/task/brain-tumor-segmentation","name":"Brain Tumor Segmentation","papers":1},{"task":"/task/ct-reconstruction","name":"CT Reconstruction","papers":1},{"task":"/task/classification-1","name":"Classification","papers":1},{"task":"/task/colorization","name":"Colorization","papers":1},{"task":"/task/computed-tomography-ct","name":"Computed Tomography (CT)","papers":1},{"task":"/task/contrastive-learning","name":"Contrastive Learning","papers":1},{"task":"/task/deep-attention","name":"Deep Attention","papers":1}],"tasks_shown":20,"n_tasks":51,"usage_by_year":[{"year":"2021","papers":3},{"year":"2022","papers":8},{"year":"2023","papers":7},{"year":"2024","papers":10},{"year":"2025","papers":3}],"row_source":"methods_table","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/coordinate-attention"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}