Mitigate Domain Shift by Primary-Auxiliary Objectives Association for Generalizing Person ReID

TL;DR

This paper proposes a novel training framework that combines primary person re-identification with auxiliary pedestrian saliency detection to improve domain generalization, using a gradient calibration mechanism for better multi-task learning.

Contribution

It introduces the PAOA mechanism to harmonize multi-task learning and extends it with PAOA+ for test-time adaptation, enhancing generalization in unseen domains.

Findings

PAOA outperforms baseline models in domain generalization tasks.

PAOA+ achieves superior results with test-time optimization.

The method effectively reduces domain bias in person ReID.

Abstract

While deep learning has significantly improved ReID model accuracy under the independent and identical distribution (IID) assumption, it has also become clear that such models degrade notably when applied to an unseen novel domain due to unpredictable/unknown domain shift. Contemporary domain generalization (DG) ReID models struggle in learning domain-invariant representation solely through training on an instance classification objective. We consider that a deep learning model is heavily influenced and therefore biased towards domain-specific characteristics, e.g., background clutter, scale and viewpoint variations, limiting the generalizability of the learned model, and hypothesize that the pedestrians are domain invariant owning they share the same structural characteristics. To enable the ReID model to be less domain-specific from these pure pedestrians, we introduce a method that guides model learning of the primary ReID instance classification objective by a concurrent auxiliary learning objective on weakly labeled pedestrian saliency detection. To solve the problem of conflicting optimization criteria in the model parameter space between the two learning objectives, we introduce a Primary-Auxiliary Objectives Association (PAOA) mechanism to calibrate the loss gradients of the auxiliary task towards the primary learning task gradients. Benefiting from the harmonious multitask learning design, our model can be extended with the recent test-time diagram to form the PAOA+, which performs on-the-fly optimization against the auxiliary objective in order to maximize the model's generative capacity in the test target domain. Experiments demonstrate the superiority of the proposed PAOA model.