Modality-Adaptive Mixup and Invariant Decomposition for RGB-Infrared Person Re-identification

RGB-infrared person re-identification is an emerging cross-modality task, which very challenging due to significant modality discrepancy between RGB and infrared images. In this work, we propose a novel modality-adaptive mixup invariant decomposition (MID) approach for towards learning modality-invariant discriminative representations. MID designs scheme generate suitable mixed images mitigating the inherent at pixel-level. It formulates procedure as Markov decision process, where actor-critic agent learns dynamical local linear interpolation policy different regions of under deep reinforcement framework. Such guarantees modality-invariance in more continuous latent space avoids manifold intrusion by corrupted samples. Moreover, further counter enforce visual semantics feature-level, employs convolution disassemble regular layer into modality-specific basis layers modality-shared coefficient layer. Extensive experimental results on two benchmarks demonstrate superior performance over state-of-the-art methods.