{"url":"/method/self-calibrated-convolutions","slug":"self-calibrated-convolutions","name":"Self-Calibrated Convolutions","full_name":"Self-Calibrated Convolutions","full_name_withheld":false,"description_markdown":"Liu et al. presented self-calibrated convolution as a means to enlarge the receptive field at each spatial location. \r\n\r\nSelf-calibrated convolution is used together with a standard convolution. It first divides the input feature $X$ into $X_{1}$ and $X_{2}$ in the channel domain. The self-calibrated convolution first uses average pooling to reduce the input size and enlarge the receptive field:\r\n\\begin{align}\r\nT_{1} = AvgPool_{r}(X_{1}) \r\n\\end{align}\r\nwhere $r$ is the filter size and stride. Then a convolution is used to model the channel relationship and a bilinear interpolation operator $Up$ is used to upsample the feature map: \r\n\r\n\\begin{align}\r\nX'_{1} = \\text{Up}(Conv_2(T_1))\r\n\\end{align}\r\n\r\nNext, element-wise multiplication finishes the self-calibrated process:\r\n\r\n\\begin{align}\r\nY'_{1} = Conv_3(X_1) \\sigma(X_1 + X'_1)\r\n\\end{align}\r\n\r\nFinally, the output feature map of is formed:\r\n\\begin{align}\r\nY_{1} &= Conv_4(Y'_{1})\r\n\\end{align}\r\n\\begin{align}\r\nY_2 &= Conv_1(X_2)\r\n\\end{align}\r\n\\begin{align}\r\nY &= [Y_1; Y_2]\r\n\\end{align}\r\nSuch self-calibrated convolution can enlarge the receptive field of a network and improve its adaptability. It achieves excellent results in image classification and certain downstream tasks such as instance segmentation, object detection and keypoint detection.","description_state":"present","introduced_year":null,"introduced_by":{"title":"Improving Convolutional Networks With Self-Calibrated Convolutions","paper":"/paper/improving-convolutional-networks-with-self","first_author":"Jiang-Jiang Liu","n_authors":5,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/improving-convolutional-networks-with-self"},"source":{"url":"http://openaccess.thecvf.com/content_CVPR_2020/html/Liu_Improving_Convolutional_Networks_With_Self-Calibrated_Convolutions_CVPR_2020_paper.html","title":"Improving Convolutional Networks With Self-Calibrated Convolutions","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":2,"archive_num_papers":2,"papers_newest_first":[{"paper":null,"title":"Self-calibrated convolution towards glioma segmentation","date":"2024-02-07","arxiv_id":"2402.05218","n_code_links":0,"syntology":null},{"paper":"/paper/improving-convolutional-networks-with-self","title":"Improving Convolutional Networks With Self-Calibrated Convolutions","date":"2020-06-01","arxiv_id":null,"n_code_links":3,"syntology":null}],"papers_shown":2,"tasks":[{"task":"/task/brain-tumor-segmentation","name":"Brain Tumor Segmentation","papers":1},{"task":"/task/instance-segmentation","name":"Instance Segmentation","papers":1},{"task":"/task/keypoint-detection","name":"Keypoint Detection","papers":1},{"task":"/task/object-detection","name":"Object Detection","papers":1},{"task":"/task/representation-learning","name":"Representation Learning","papers":1},{"task":"/task/segmentation","name":"Segmentation","papers":1},{"task":"/task/semantic-segmentation","name":"Semantic Segmentation","papers":1},{"task":"/task/tumor-segmentation","name":"Tumor Segmentation","papers":1},{"task":"/task/object-detection-1","name":"object-detection","papers":1}],"tasks_shown":9,"n_tasks":9,"usage_by_year":[{"year":"2020","papers":1},{"year":"2024","papers":1}],"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/self-calibrated-convolutions"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}