深入探究对比语言-图像预训练的可解释性
A Closer Look at the Explainability of Contrastive Language-Image Pre-training
AI总结：
针对CLIP可解释性差、易关注背景且存在噪声激活的问题，提出无需微调的CLIP Surgery方法，通过修改推理架构与特征提升可解释性，还支持多模态可视化与开放词汇任务拓展。
摘要：
对比语言-图像预训练（CLIP）是一种功能强大的视觉-语言模型，已在各类任务中展现出显著优势。但我们发现其可解释性存在若干问题，这会削弱模型可信度，也限制了相关任务的性能。具体而言，我们发现CLIP倾向于关注背景区域而非前景，可视化结果中无关位置存在噪声激活。这些现象与基于类注意力图（CAM）的传统可解释性方法的表现相悖——传统方法中原始模型无需对齐，仅通过全局监督就能突出局部前景区域。为解决这些问题，我们深入分析了其架构与特征。经全面研究发现，原始自注意力关联的语义区域不一致，导致可视化结果出现反向偏差；而噪声激活则源于类别间的冗余特征。基于这些发现，我们提出了用于生成可靠CAM的CLIP Surgery方法，该方法无需像经典CAM方法那样进行额外微调，就能对推理架构和特征进行类似“手术”的修改。该方法大幅提升了CLIP的可解释性，效果远超现有方法。此外，它还支持多模态可视化，在无需额外对齐的情况下拓展了原始CLIP在开放词汇任务上的能力。代码可在https://github.com/xmed-lab/CLIP_Surgery获取。
英文摘要：
Contrastive language-image pre-training (CLIP) is a powerful vision-language model that has shown great benefits for various tasks. However, we have identified some issues with its explainability, which undermine its credibility and limit the capacity for related tasks. Specifically, we find that CLIP tends to focus on background regions rather than foregrounds, with noisy activations at irrelevant positions on the visualization results. These phenomena conflict with conventional explainability methods based on the class attention map (CAM), where the raw model can highlight the local foreground regions using global supervision without alignment. To address these problems, we take a closer look at its architecture and features. Based on thorough analyses, we find the raw self-attentions link to inconsistent semantic regions, resulting in the opposite visualization. Besides, the noisy activations are owing to redundant features among categories. Building on these insights, we propose the CLIP Surgery for reliable CAM, a method that allows surgery-like modifications to the inference architecture and features, without further fine-tuning as classical CAM methods. This approach significantly improves the explainability of CLIP, surpassing existing methods by large margins. Besides, it enables multimodal visualization and extends the capacity of raw CLIP on open-vocabulary tasks without extra alignment. The code is available at https://github.com/xmed-lab/CLIP_Surgery.
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。
