Description
Self-Attention Network (SANet) proposes two variations of self-attention used for image recognition: 1) pairwise self-attention which generalizes standard dot-product attention and is fundamentally a set operator, and 2) patchwise self-attention which is strictly more powerful than convolution.
Papers Using This Method
Counting Manatee Aggregations using Deep Neural Networks and Anisotropic Gaussian Kernel2023-11-04Spatial-Assistant Encoder-Decoder Network for Real Time Semantic Segmentation2023-09-19Strip Attention for Image Restoration2023-08-01Structure Aggregation for Cross-Spectral Stereo Image Guided Denoising2023-01-01Artistic Arbitrary Style Transfer2022-12-21Playing Lottery Tickets in Style Transfer Models2022-03-25Exploring Separable Attention for Multi-Contrast MR Image Super-Resolution2021-09-03Shallow Attention Network for Polyp Segmentation2021-08-02Learning Camera Localization via Dense Scene Matching2021-03-31Scale-aware Neural Network for Semantic Segmentation of Multi-resolution Remote Sensing Images2021-03-14Exploring Self-attention for Image Recognition2020-04-28