Adversarial Pairwise Reverse Attention for Camera Performance Imbalance in Person Re-identification: New Dataset and Metrics
This addresses a long-ignored issue in Person ReID that affects system robustness due to uneven camera data distributions and properties, though it is incremental as it builds on existing ReID methods.
The paper tackles the problem of camera performance imbalance in Person Re-Identification (Person ReID) by collecting a real-world dataset from 38 cameras and proposing new metrics to quantify this imbalance, resulting in the development of the Adversarial Pairwise Reverse Attention (APRA) Module to learn camera-invariant features.
Existing evaluation metrics for Person Re-Identification (Person ReID) models focus on system-wide performance. However, our studies reveal weaknesses due to the uneven data distributions among cameras and different camera properties that expose the ReID system to exploitation. In this work, we raise the long-ignored ReID problem of camera performance imbalance and collect a real-world privacy-aware dataset from 38 cameras to assist the study of the imbalance issue. We propose new metrics to quantify camera performance imbalance and further propose the Adversarial Pairwise Reverse Attention (APRA) Module to guide the model learning the camera invariant feature with a novel pairwise attention inversion mechanism.