Dynamic Textual Prompt For Rehearsal-free Lifelong Person Re-identification

TL;DR

This paper introduces a dynamic textual prompt framework for lifelong person re-identification that avoids data retention, mitigates catastrophic forgetting, and enhances cross-domain invariance by leveraging shared textual descriptions.

Contribution

It proposes a novel task-driven dynamic textual prompt method with modules for prompt fusion, feature alignment, and knowledge distillation to improve lifelong ReID without sample retention.

Findings

Outperforms state-of-the-art methods in various settings

Effectively mitigates catastrophic forgetting

Achieves better cross-domain invariance

Abstract

Lifelong person re-identification attempts to recognize people across cameras and integrate new knowledge from continuous data streams. Key challenges involve addressing catastrophic forgetting caused by parameter updating and domain shift, and maintaining performance in seen and unseen domains. Many previous works rely on data memories to retain prior samples. However, the amount of retained data increases linearly with the number of training domains, leading to continually increasing memory consumption. Additionally, these methods may suffer significant performance degradation when data preservation is prohibited due to privacy concerns. To address these limitations, we propose using textual descriptions as guidance to encourage the ReID model to learn cross-domain invariant features without retaining samples. The key insight is that natural language can describe pedestrian instances with an invariant style, suggesting a shared textual space for any pedestrian images. By leveraging this shared textual space as an anchor, we can prompt the ReID model to embed images from various domains into a unified semantic space, thereby alleviating catastrophic forgetting caused by domain shifts. To achieve this, we introduce a task-driven dynamic textual prompt framework in this paper. This model features a dynamic prompt fusion module, which adaptively constructs and fuses two different textual prompts as anchors. This effectively guides the ReID model to embed images into a unified semantic space. Additionally, we design a text-visual feature alignment module to learn a more precise mapping between fine-grained visual and textual features. We also developed a learnable knowledge distillation module that allows our model to dynamically balance retaining existing knowledge with acquiring new knowledge. Extensive experiments demonstrate that our method outperforms SOTAs under various settings.