{"url":"/method/gct","slug":"gct","name":"GCT","full_name":"Gated Channel Transformation","full_name_withheld":false,"description_markdown":"GCT first collects global information by computing the l2-norm of each channel. Next, a learnable vector $ \\alpha $ is applied to scale the feature. Then a competition mechanism is adopted by channel normalization to interact between channels.\r\n\r\nUnlike previous methods, GCT first collects global information by computing the $l_{2}$-norm of each channel. \r\nNext, a learnable vector $\\alpha$ is applied to scale the feature.\r\nThen a competition mechanism is adopted by \r\nchannel normalization to interact between channels. \r\nLike other common normalization methods, \r\na learnable scale parameter $\\gamma$ and bias $\\beta$ are applied to \r\nrescale the normalization.\r\nHowever, unlike previous methods,\r\nGCT adopts tanh activation to control the attention vector.\r\nFinally, it not only multiplies the input by the attention vector but also adds an identity connection. GCT can be written as: \r\n\\begin{align}\r\n    s = F_\\text{gct}(X, \\theta) & = \\tanh (\\gamma CN(\\alpha \\text{Norm}(X)) + \\beta)\r\n\\end{align}\r\n\\begin{align}\r\n    Y & = s  X + X\r\n\\end{align}\r\n\r\nwhere $\\alpha$, $\\beta$ and $\\gamma$ are trainable parameters. $\\text{Norm}(\\cdot)$ indicates the $L2$-norm of each channel. $CN$ is  channel normalization.\r\n\r\nA GCT block has fewer parameters than an SE block, and as it is  lightweight, \r\n can be added after each convolutional layer of a CNN.","description_state":"present","introduced_year":null,"introduced_by":{"title":"Gated Channel Transformation for Visual Recognition","paper":"/paper/gated-channel-transformation-for-visual","first_author":"Zongxin Yang","n_authors":4,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/gated-channel-transformation-for-visual"},"source":{"url":"https://arxiv.org/abs/1909.11519v2","title":"Gated Channel Transformation for Visual Recognition","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":8,"archive_num_papers":8,"papers_newest_first":[{"paper":null,"title":"STEAM: Squeeze and Transform Enhanced Attention Module","date":"2024-12-12","arxiv_id":"2412.09023","n_code_links":0,"syntology":null},{"paper":"/paper/a-generative-framework-to-bridge-data-driven","title":"Generative causal testing to bridge data-driven models and scientific theories in language neuroscience","date":"2024-10-01","arxiv_id":"2410.00812","n_code_links":1,"syntology":{"ran":8,"of":11,"unverified":3,"pointer_only":0}},{"paper":"/paper/teltrans-applying-multi-type-telecom-data-to","title":"TelTrans: Applying Multi-Type Telecom Data to Transportation Evaluation and Prediction via Multifaceted Graph Modeling","date":"2024-01-06","arxiv_id":"2401.03138","n_code_links":0,"syntology":{"ran":2,"of":7,"unverified":5,"pointer_only":7}},{"paper":null,"title":"Granger Causality for Predictability in Dynamic Mode Decomposition","date":"2022-10-23","arxiv_id":"2210.12737","n_code_links":0,"syntology":null},{"paper":null,"title":"Generalised Co-Salient Object Detection","date":"2022-08-20","arxiv_id":"2208.09668","n_code_links":0,"syntology":null},{"paper":null,"title":"GCT: Graph Co-Training for Semi-Supervised Few-Shot Learning","date":"2022-03-15","arxiv_id":"2203.07738","n_code_links":0,"syntology":null},{"paper":null,"title":"Integrating Fréchet distance and AI reveals the evolutionary trajectory and origin of SARS-CoV-2","date":"2021-10-14","arxiv_id":"2110.07696","n_code_links":0,"syntology":null},{"paper":"/paper/gated-channel-transformation-for-visual","title":"Gated Channel Transformation for Visual Recognition","date":"2019-09-25","arxiv_id":"1909.11519","n_code_links":3,"syntology":null}],"papers_shown":8,"tasks":[{"task":"/task/object-detection","name":"Object Detection","papers":3},{"task":"/task/object-detection-1","name":"object-detection","papers":3},{"task":"/task/image-classification","name":"Image Classification","papers":2},{"task":"/task/instance-segmentation","name":"Instance Segmentation","papers":2},{"task":"/task/semantic-segmentation","name":"Semantic Segmentation","papers":2},{"task":"/task/image-classification","name":"image-classification","papers":2},{"task":"/task/classification-1","name":"Classification","papers":1},{"task":"/task/co-saliency-detection","name":"Co-Salient Object Detection","papers":1},{"task":"/task/few-shot-learning","name":"Few-Shot Learning","papers":1},{"task":"/task/classification","name":"General Classification","papers":1},{"task":"/task/graph-learning","name":"Graph Learning","papers":1},{"task":"/task/graph-neural-network","name":"Graph Neural Network","papers":1},{"task":"/task/object","name":"Object","papers":1},{"task":"/task/prediction","name":"Prediction","papers":1},{"task":"/task/salient-object-detection-1","name":"Salient Object Detection","papers":1},{"task":"/task/self-supervised-learning","name":"Self-Supervised Learning","papers":1},{"task":"/task/time-series","name":"Time Series Analysis","papers":1},{"task":"/task/traffic-prediction","name":"Traffic Prediction","papers":1},{"task":"/task/video-classification","name":"Video Classification","papers":1}],"tasks_shown":19,"n_tasks":19,"usage_by_year":[{"year":"2019","papers":1},{"year":"2021","papers":1},{"year":"2022","papers":3},{"year":"2024","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/gct"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}