CVPR 2025 | TAB:Token聚合块,仅训练阶段更新聚类中心,推理速度翻倍!
论文: CATANet: Efficient Content-Aware Token Aggregation for Lightweight Image Super-Resolution 作者: Xin Liu, Jie Liu, Jie Tang, Gangshan Wu(State Key Lab for Novel Software Technology, Nanjing University) 发表: CVPR 2025 论文链接: arXiv:2503.06896 代码…