Learning Visual Proxy for Compositional Zero-Shot Learning
This work improves compositional generalization for AI systems in visual recognition tasks, representing an incremental advancement over existing methods.
The paper tackles the problem of recognizing novel attribute-object compositions in Compositional Zero-Shot Learning by addressing modality gaps and lack of fine-grained visual cues, achieving state-of-the-art performance in closed-world scenarios and competitive results in open-world settings across four benchmarks.
Compositional Zero-Shot Learning (CZSL) aims to recognize novel attribute-object compositions by leveraging knowledge from seen compositions. Current methods align textual prototypes with visual features via Vision-Language Models (VLMs), but suffer from two limitations: (1) modality gaps hinder the discrimination of semantically similar pairs, and (2) single-modal textual prototypes lack fine-grained visual cues. In this paper, we introduce Visual Proxy Learning, a method that reduces modality gaps and enhances compositional generalization. We initialize visual proxies for attributes, objects, and their compositions using text representations and optimize the visual space to capture fine-grained cues, improving visual representations. Additionally, we propose Cross-Modal Joint Learning (CMJL), which imposes cross-modal constraints between the text-image and fine-grained visual spaces, improving generalization for unseen compositions and discriminating similar pairs. Experiments show state-of-the-art performance in closed-world scenarios and competitive results in open-world settings across four CZSL benchmarks, demonstrating the effectiveness of our approach in compositional generalization.