BiCnet-TKS: Learning Efficient Spatial-Temporal Representation for Video Person Re-Identification

TL;DR

This paper introduces BiCnet-TKS, a novel efficient model for video person re-identification that combines spatial complementarity and adaptive temporal relation modeling, achieving superior accuracy with reduced computation.

Contribution

The paper proposes a Bilateral Complementary Network with a Temporal Kernel Selection block for improved spatial-temporal feature extraction in video reID, with enhanced efficiency.

Findings

Outperforms state-of-the-art methods on multiple benchmarks.

Reduces computational cost by about 50%.

Effectively captures both detailed and contextual features.

Abstract

In this paper, we present an efficient spatial-temporal representation for video person re-identification (reID). Firstly, we propose a Bilateral Complementary Network (BiCnet) for spatial complementarity modeling. Specifically, BiCnet contains two branches. Detail Branch processes frames at original resolution to preserve the detailed visual clues, and Context Branch with a down-sampling strategy is employed to capture long-range contexts. On each branch, BiCnet appends multiple parallel and diverse attention modules to discover divergent body parts for consecutive frames, so as to obtain an integral characteristic of target identity. Furthermore, a Temporal Kernel Selection (TKS) block is designed to capture short-term as well as long-term temporal relations by an adaptive mode. TKS can be inserted into BiCnet at any depth to construct BiCnetTKS for spatial-temporal modeling. Experimental results on multiple benchmarks show that BiCnet-TKS outperforms state-of-the-arts with about 50% less computations. The source code is available at https://github.com/ blue-blue272/BiCnet-TKS.