CO-EVO: Co-evolving Semantic Anchoring and Style Diversification for Federated DG-ReID

TL;DR

CO-EVO introduces a co-evolutionary federated learning framework for person re-identification that enhances cross-domain generalization by combining semantic anchoring and style diversification.

Contribution

It proposes a novel co-evolutionary mechanism that jointly learns domain-agnostic semantic anchors and synthesizes style variations to improve federated DG-ReID.

Findings

Achieves state-of-the-art performance on federated DG-ReID benchmarks.

Effectively filters out camera biases to focus on universal identity features.

Enhances robustness through realistic style perturbations.

Abstract

Federated domain generalization for person re-identification (FedDG-ReID) aims to collaboratively train a pedestrian retrieval model across multiple decentralized source domains such that it can generalize to unseen target environments without compromising raw data privacy. However, this task is significantly challenged by the inherent stylistic gaps across decentralized clients. Without global supervision, models easily succumb to shortcut learning where representations overfit to domain specific camera biases rather than universal identity features. We propose CO-EVO, a novel federated framework that resolves this semantic-style conflict through a co-evolutionary mechanism. On the semantic side, Camera-Invariant Semantic Anchoring (CSA) learns identity prompts with cross-camera consistency to establish purified and domain-agnostic anchors that filter out local imaging noise. On the visual side, Global Style Diversification (GSD), powered by a Global Camera-Style Bank (GCSB), synthesizes realistic perturbations to expand the visual boundaries of training data. The core of CO-EVO is its co-evolutionary loop where purified anchors act as gravitational centers to guide the image encoder toward robust anatomical attributes amidst diverse style variations. Extensive experiments demonstrate that CO-EVO achieves state-of-the-art (SOTA) performance, proving that the synergy between semantic purification and style expansion is essential for robust cross-domain generalization. Our code is available at: https://github.com/NanYiyuzurn/ACL-LGPS-2026.