Feature-Distribution Perturbation and Calibration for Generalized Person ReID

TL;DR

This paper introduces PECA, a novel method for person ReID that enhances model generalization across unseen domains by perturbing and calibrating feature distributions to reduce domain bias and overfitting.

Contribution

The paper proposes a feature-distribution perturbation and calibration framework that improves person ReID performance across multiple unseen domains, addressing domain bias and overfitting issues.

Findings

PECA outperforms state-of-the-art methods on eight datasets.

The method effectively reduces domain bias and improves generalization.

Extensive experiments validate the robustness of PECA across diverse scenarios.

Abstract

Person Re-identification (ReID) has been advanced remarkably over the last 10 years along with the rapid development of deep learning for visual recognition. However, the i.i.d. (independent and identically distributed) assumption commonly held in most deep learning models is somewhat non-applicable to ReID considering its objective to identify images of the same pedestrian across cameras at different locations often of variable and independent domain characteristics that are also subject to view-biased data distribution. In this work, we propose a Feature-Distribution Perturbation and Calibration (PECA) method to derive generic feature representations for person ReID, which is not only discriminative across cameras but also agnostic and deployable to arbitrary unseen target domains. Specifically, we perform per-domain feature-distribution perturbation to refrain the model from overfitting to the domain-biased distribution of each source (seen) domain by enforcing feature invariance to distribution shifts caused by perturbation. Furthermore, we design a global calibration mechanism to align feature distributions across all the source domains to improve the model generalization capacity by eliminating domain bias. These local perturbation and global calibration are conducted simultaneously, which share the same principle to avoid models overfitting by regularization respectively on the perturbed and the original distributions. Extensive experiments were conducted on eight person ReID datasets and the proposed PECA model outperformed the state-of-the-art competitors by significant margins.