Text-Phase Synergy Network with Dual Priors for Unsupervised Cross-Domain Image Retrieval

TL;DR

This paper introduces TPSNet, a novel unsupervised cross-domain image retrieval method that uses dual priors from CLIP-generated text prompts and domain-invariant phase features to enhance semantic guidance and domain alignment.

Contribution

The paper proposes TPSNet, integrating text and phase priors to improve unsupervised cross-domain image retrieval beyond existing pseudo-label based methods.

Findings

TPSNet outperforms state-of-the-art methods on UCDIR benchmarks.

Using CLIP-based class prompts provides more accurate semantic supervision.

Domain-invariant phase features effectively bridge domain gaps while preserving semantics.

Abstract

This paper studies unsupervised cross-domain image retrieval (UCDIR), which aims to retrieve images of the same category across different domains without relying on labeled data. Existing methods typically utilize pseudo-labels, derived from clustering algorithms, as supervisory signals for intra-domain representation learning and cross-domain feature alignment. However, these discrete pseudo-labels often fail to provide accurate and comprehensive semantic guidance. Moreover, the alignment process frequently overlooks the entanglement between domain-specific and semantic information, leading to semantic degradation in the learned representations and ultimately impairing retrieval performance. This paper addresses the limitations by proposing a Text-Phase Synergy Network with Dual Priors(TPSNet). Specifically, we first employ CLIP to generate a set of class-specific prompts per domain, termed as domain prompt, serving as a text prior that offers more precise semantic supervision. In parallel, we further introduce a phase prior, represented by domain-invariant phase features, which is integrated into the original image representations to bridge the domain distribution gaps while preserving semantic integrity. Leveraging the synergy of these dual priors, TPSNet significantly outperforms state-of-the-art methods on UCDIR benchmarks.