面向高效视频无监督域自适应的可学习运动聚焦 Token 化
Learnable Motion-Focused Tokenization for Effective and Efficient Video Unsupervised Domain Adaptation
把视频帧切成 patch token 后学习丢弃低运动、冗余的背景 token，只保留运动丰富的动作相关 token 再做域自适应——这个可学习 token 选择模块可以直接插到任何视频 Transformer 的域自适应或高效推理流程里，用来同时压掉背景域偏移和计算量。
原文依据 learns to discard low-motion, redundant tokens, primarily corresponding to background regions, while retaining motion-rich, action-relevant tokens
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论视频无监督域自适应（VUDA）在动作识别中是一项重大挑战，需要将模型从有标注源域适配到无标注目标域。尽管近期取得了进展，现有 VUDA 方法往往达不到全监督性能，一个关键原因是普遍存在的静态且无信息的背景加剧了域偏移。此外，先前方法在很大程度上忽视了计算效率，限制了实际应用。为解决这些问题，我们提出了面向 VUDA 的可学习运动聚焦 Token 化（LMFT）。LMFT 将视频帧 token 化为图像块 token，并学习丢弃主要对应背景区域的低运动、冗余 token，同时保留运动丰富、动作相关的 token 用于适配。在三个标准 VUDA 基准、21 种域自适应设置上的大量实验表明，采用 LMFT 的 VUDA 框架在显著降低计算开销的同时达到了最优性能。因此，LMFT 使 VUDA 兼具有效性与计算高效性。
英文原文摘要 1102 字符
Video Unsupervised Domain Adaptation (VUDA) poses a significant challenge in action recognition, requiring the adaptation of a model from a labeled source domain to an unlabeled target domain. Despite recent advances, existing VUDA methods often fall short of fully supervised performance, a key reason being the prevalence of static and uninformative backgrounds that exacerbate domain shifts. Additionally, prior approaches largely overlook computational efficiency, limiting real-world adoption. To address these issues, we propose Learnable Motion-Focused Tokenization (LMFT) for VUDA. LMFT tokenizes video frames into patch tokens and learns to discard low-motion, redundant tokens, primarily corresponding to background regions, while retaining motion-rich, action-relevant tokens for adaptation. Extensive experiments on three standard VUDA benchmarks across 21 domain adaptation settings show that our VUDA framework with LMFT achieves state-of-the-art performance while significantly reducing computational overhead. LMFT thus enables VUDA that is both effective and computationally efficient.