arXiv 2604.221902026-04-24cs.CVcs.AI

从全局到局部：重新思考CLIP特征聚合用于人重识别

From Global to Local: Rethinking CLIP Feature Aggregation for Person Re-Identification

Department of Electrical and Computer Engineering, University of Washington（华盛顿大学电气与计算机工程系）；Applied Physics Laboratory, University of Washington（华盛顿大学应用物理实验室）

Aotian Zheng, Winston Sun, Bahaa Alattar, Vitaly Ablavsky, Jenq-Neng Hwang

备注：: 14 pages, 7 figures

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出SAGA-ReID，通过在CLIP文本嵌入空间中对齐中间补丁标记与锚向量，改进CLIP基于的人重识别方法，提升在遮挡和跨摄像头变化下的鲁棒性，实验表明其在遮挡增加时表现更优，达到+10.6 Rank-1的提升。

摘要：

基于CLIP的人重识别方法通过将空间特征聚合到单一全局[CLS]标记中，优化图像-文本对齐而非空间选择性，导致在遮挡和跨摄像头变化下的表示脆弱。我们提出SAGA-ReID，通过在CLIP文本嵌入空间中对齐中间补丁标记与锚向量来重建身份表示，强调空间稳定的证据同时抑制损坏或缺失的区域，无需对单个图像进行文本描述。受控实验在两种质上不同的条件下隔离聚合机制——合成遮挡，其中身份信号缺失，以及现实中的人类干扰，其中重叠的人引入语义混淆信号——SAGA在遮挡增加时相对于全局池化在两种条件下均显著占优。基准评估证实了在标准和遮挡设置中相对于CLIP-ReID的一致提升，最大的改进出现在全局池化最不可靠的地方：在遮挡基准上达到+10.6 Rank-1。SAGA的聚合在更强的主干网络上优于专门的序列补丁聚合，证实了结构化重建解决了主干质量和架构复杂性本身无法解决的瓶颈。代码可在https://github.com/ipl-uw/Structured-Anchor-Guided-Aggregation-for-ReID获取。

英文摘要：

CLIP-based person re-identification (ReID) methods aggregate spatial features into a single global \texttt{[CLS]} token optimized for image-text alignment rather than spatial selectivity, making representations fragile under occlusion and cross-camera variation. We propose SAGA-ReID, which reconstructs identity representations by aligning intermediate patch tokens with anchor vectors parameterized in CLIP's text embedding space -- emphasizing spatially stable evidence while suppressing corrupted or absent regions, without requiring textual descriptions of individual images. Controlled experiments isolate the aggregation mechanism under two qualitatively distinct conditions -- synthetic masking, where identity signal is absent, and realistic human distractors, where an overlapping person introduces semantically confusing signal -- with SAGA's advantage over global pooling growing substantially as occlusion increases across both conditions. Benchmark evaluations confirm consistent gains over CLIP-ReID across standard and occluded settings, with the largest improvements where global pooling is most unreliable: up to +10.6 Rank-1 on occluded benchmarks. SAGA's aggregation outperforms dedicated sequential patch aggregation on a stronger backbone, confirming that structured reconstruction addresses a bottleneck that backbone quality and architectural complexity alone cannot resolve. Code available at https://github.com/ipl-uw/Structured-Anchor-Guided-Aggregation-for-ReID.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。