VLM-KD：面向长尾视觉识别的视觉语言模型知识蒸馏
VLM-KD: Knowledge Distillation from VLM for Long-Tail Visual Recognition
AI总结：
本文提出 VLM-KD 框架，利用现成视觉语言模型生成新颖文本监督并蒸馏到随机初始化的视觉编码器中，在多个长尾视觉识别基准上超越现有最先进方法。
摘要：
在视觉识别中，知识蒸馏通常涉及将知识从大型、训练有素的教师模型迁移到较小的学生模型。本文提出了一种有效的方法，从现成的视觉语言模型（VLM）中蒸馏知识，证明其除了来自传统纯视觉教师模型的监督之外，还能提供新颖的监督信号。我们的关键技术贡献是开发了一个框架，该框架生成新颖的文本监督，并将自由形式的文本蒸馏到视觉编码器中。我们在多个基准数据集上展示了所提方法（称为 VLM-KD）的有效性，表明其超越了若干最先进的长尾视觉分类器。据我们所知，这项工作是首个利用现成 VLM 生成的文本监督进行知识蒸馏，并将其应用于普通的随机初始化视觉编码器的工作。
英文摘要：
For visual recognition, knowledge distillation typically involves transferring knowledge from a large, well-trained teacher model to a smaller student model. In this paper, we introduce an effective method to distill knowledge from an off-the-shelf vision-language model (VLM), demonstrating that it provides novel supervision in addition to those from a conventional vision-only teacher model. Our key technical contribution is the development of a framework that generates novel text supervision and distills free-form text into a vision encoder. We showcase the effectiveness of our approach, termed VLM-KD, across various benchmark datasets, showing that it surpasses several state-of-the-art long-tail visual classifiers. To our knowledge, this work is the first to utilize knowledge distillation with text supervision generated by an off-the-shelf VLM and apply it to vanilla randomly initialized vision encoders.
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。
