ComKD-CLIP: Comprehensive Knowledge Distillation for Contrastive Language-Image Pre-traning Model

TL;DR

This paper introduces ComKD-CLIP, a novel knowledge distillation approach that effectively transfers knowledge from large CLIP models to smaller ones, maintaining high performance while reducing computational requirements.

Contribution

The paper proposes a comprehensive distillation framework with two mechanisms, IFAlign and EduAttention, to improve smaller CLIP models' performance by mimicking large teacher models.

Findings

Outperforms existing methods on 11 datasets

Achieves comparable performance with significantly fewer parameters

Demonstrates effective knowledge transfer in multimodal tasks

Abstract

Contrastive Language-Image Pre-training (CLIP) models excel in integrating semantic information between images and text through contrastive learning techniques. It has achieved remarkable performance in various multimodal tasks. However, the deployment of large CLIP models is hindered in resource-limited environments, while smaller models frequently fail to meet the performance benchmarks required for practical applications. In this paper, we propose a novel approach, ComKD-CLIP: Comprehensive Knowledge Distillation for Contrastive Language-Image Pre-traning Model, which aims to comprehensively distill the knowledge from a large teacher CLIP model into a smaller student model, ensuring comparable performance with significantly reduced parameters. ComKD-CLIP is composed of two key mechanisms: Image Feature Alignment (IFAlign) and Educational Attention (EduAttention). IFAlign makes the image features extracted by the student model closely match those extracted by the teacher model, enabling the student to learn teacher's knowledge of extracting image features. EduAttention explores the cross-relationships between text features extracted by the teacher model and image features extracted by the student model, enabling the student model to learn how the teacher model integrates text-image features. In addition, ComKD-CLIP can refine the knowledge distilled from IFAlign and EduAttention by leveraging the text-image feature fusion results of the teacher model, ensuring the student model accurately absorbs the teacher's knowledge. Extensive experiments conducted on 11 datasets have demonstrated the superiority of the proposed method.