KPL: Training-Free Medical Knowledge Mining of Vision-Language Models

TL;DR

This paper introduces KPL, a novel method that enhances zero-shot medical image classification by mining and leveraging knowledge from CLIP, addressing class representation and modal gap challenges.

Contribution

KPL is a training-free approach that enriches semantic proxies with knowledge descriptions and uses multimodal proxy learning to improve medical image classification performance.

Findings

KPL outperforms baseline methods on medical datasets.

KPL demonstrates stability and effectiveness in zero-shot classification.

KPL also shows improvements on natural image datasets.

Abstract

Visual Language Models such as CLIP excel in image recognition due to extensive image-text pre-training. However, applying the CLIP inference in zero-shot classification, particularly for medical image diagnosis, faces challenges due to: 1) the inadequacy of representing image classes solely with single category names; 2) the modal gap between the visual and text spaces generated by CLIP encoders. Despite attempts to enrich disease descriptions with large language models, the lack of class-specific knowledge often leads to poor performance. In addition, empirical evidence suggests that existing proxy learning methods for zero-shot image classification on natural image datasets exhibit instability when applied to medical datasets. To tackle these challenges, we introduce the Knowledge Proxy Learning (KPL) to mine knowledge from CLIP. KPL is designed to leverage CLIP's multimodal understandings for medical image classification through Text Proxy Optimization and Multimodal Proxy Learning. Specifically, KPL retrieves image-relevant knowledge descriptions from the constructed knowledge-enhanced base to enrich semantic text proxies. It then harnesses input images and these descriptions, encoded via CLIP, to stably generate multimodal proxies that boost the zero-shot classification performance. Extensive experiments conducted on both medical and natural image datasets demonstrate that KPL enables effective zero-shot image classification, outperforming all baselines. These findings highlight the great potential in this paradigm of mining knowledge from CLIP for medical image classification and broader areas.