UCDR-Adapter: Exploring Adaptation of Pre-Trained Vision-Language Models for Universal Cross-Domain Retrieval

TL;DR

UCDR-Adapter introduces a novel method combining adapters and dynamic prompt generation to improve universal cross-domain retrieval, enabling better adaptation to unseen domains and classes without relying on semantic labels.

Contribution

It proposes a two-phase training strategy with source adapter learning and target prompt generation, enhancing pre-trained vision-language models for flexible, zero-shot retrieval across diverse domains.

Findings

Outperforms state-of-the-art methods on multiple UCDR benchmarks

Demonstrates robust generalization to unseen domains and classes

Achieves higher retrieval accuracy with dynamic prompts during inference

Abstract

Universal Cross-Domain Retrieval (UCDR) retrieves relevant images from unseen domains and classes without semantic labels, ensuring robust generalization. Existing methods commonly employ prompt tuning with pre-trained vision-language models but are inherently limited by static prompts, reducing adaptability. We propose UCDR-Adapter, which enhances pre-trained models with adapters and dynamic prompt generation through a two-phase training strategy. First, Source Adapter Learning integrates class semantics with domain-specific visual knowledge using a Learnable Textual Semantic Template and optimizes Class and Domain Prompts via momentum updates and dual loss functions for robust alignment. Second, Target Prompt Generation creates dynamic prompts by attending to masked source prompts, enabling seamless adaptation to unseen domains and classes. Unlike prior approaches, UCDR-Adapter dynamically adapts to evolving data distributions, enhancing both flexibility and generalization. During inference, only the image branch and generated prompts are used, eliminating reliance on textual inputs for highly efficient retrieval. Extensive benchmark experiments show that UCDR-Adapter consistently outperforms ProS in most cases and other state-of-the-art methods on UCDR, U(c)CDR, and U(d)CDR settings.