A Siamese Long Short-Term Memory Architecture for Human Re-Identification

TL;DR

This paper introduces a novel Siamese LSTM architecture for human re-identification that processes image regions sequentially, leveraging contextual information to improve discriminative feature representation and achieve better performance on standard datasets.

Contribution

The paper proposes a new Siamese LSTM model that captures spatial dependencies in image regions for re-identification, enhancing feature discrimination over existing methods.

Findings

Improved re-identification accuracy on Market-1501, CUHK03, and VIPeR datasets.

LSTM-based model outperforms baseline without LSTM units.

Visualization shows meaningful patterns learned by the LSTM cells.

Abstract

Matching pedestrians across multiple camera views known as human re-identification (re-identification) is a challenging problem in visual surveillance. In the existing works concentrating on feature extraction, representations are formed locally and independent of other regions. We present a novel siamese Long Short-Term Memory (LSTM) architecture that can process image regions sequentially and enhance the discriminative capability of local feature representation by leveraging contextual information. The feedback connections and internal gating mechanism of the LSTM cells enable our model to memorize the spatial dependencies and selectively propagate relevant contextual information through the network. We demonstrate improved performance compared to the baseline algorithm with no LSTM units and promising results compared to state-of-the-art methods on Market-1501, CUHK03 and VIPeR datasets. Visualization of the internal mechanism of LSTM cells shows meaningful patterns can be learned by our method.