Pre-Trained Vision-Language Models as Partial Annotators

TL;DR

This paper introduces a novel weakly-supervised learning approach using pre-trained vision-language models like CLIP to generate partial labels for image classification, improving performance without extra labeling effort.

Contribution

It proposes a collaborative label purification and self-training framework leveraging noisy partial labels from CLIP, enhancing downstream image classification performance.

Findings

Achieves significantly better results than zero-shot inference.

Outperforms existing weakly supervised and few-shot methods.

Produces smaller, efficient deployed models.

Abstract

Pre-trained vision-language models learn massive data to model unified representations of images and natural languages, which can be widely applied to downstream machine learning tasks. In addition to zero-shot inference, in order to better adapt pre-trained models to the requirements of downstream tasks, people usually use methods such as few-shot or parameter-efficient fine-tuning and knowledge distillation. However, annotating samples is laborious, while a large number of unlabeled samples can be easily obtained. In this paper, we investigate a novel "pre-trained annotating - weakly-supervised learning" paradigm for pre-trained model application and experiment on image classification tasks. Specifically, based on CLIP, we annotate image samples with multiple prompt templates to obtain multiple candidate labels to form the noisy partial label dataset, and design a collaborative consistency regularization algorithm to solve this problem. Our method simultaneously trains two neural networks, which collaboratively purify training labels for each other and obtain pseudo-labels for self-training, while adopting prototypical similarity alignment and noisy supervised contrastive learning to optimize model representation. In experiments, our method achieves performances far beyond zero-shot inference without introducing additional label information, and outperforms other weakly supervised learning and few-shot fine-tuning methods, and obtains smaller deployed models. Our code is available at: \url{https://anonymous.4open.science/r/Co-Reg-8CF9}.