Improving Zero-shot Generalization of Learned Prompts via Unsupervised Knowledge Distillation
This work addresses the need for better zero-shot generalization in vision-language models, offering a parameter-efficient method that is incremental but practical for applications with limited labeled data.
The paper tackles the problem of vision-language models underperforming in zero-shot generalization compared to supervised methods by proposing an unsupervised knowledge distillation approach for prompt learning, which improves performance on multiple benchmark datasets without requiring labeled examples.
Vision-Language Models (VLMs) demonstrate remarkable zero-shot generalization to unseen tasks, but fall short of the performance of supervised methods in generalizing to downstream tasks with limited data. Prompt learning is emerging as a parameter-efficient method for adapting VLMs, but state-of-the-art approaches require annotated samples. In this paper we propose a novel approach to prompt learning based on unsupervised knowledge distillation from more powerful models. Our approach, which we call Knowledge Distillation Prompt Learning (KDPL), can be integrated into existing prompt learning techniques and eliminates the need for labeled examples during adaptation. Our experiments on more than ten standard benchmark datasets demonstrate that KDPL is very effective at improving generalization of learned prompts for zero-shot domain generalization, zero-shot cross-dataset generalization, and zero-shot base-to-novel class generalization problems. KDPL requires no ground-truth labels for adaptation, and moreover we show that even in the absence of any knowledge of training class names it can be used to effectively transfer knowledge. The code is publicly available at https://github.com/miccunifi/KDPL.