Sparse re-id: Block sparsity for person re-identification

Abstract

This paper presents a novel approach to solve the problem of person re-identification in non-overlapping camera views. We hypothesize that the feature vector of a probe image approximately lies in the linear span of the corresponding gallery feature vectors in a learned embedding space. We then formulate the re-identification problem as a block sparse recovery problem and solve the associated optimization problem using the alternating directions framework. We evaluate our approach on the publicly available PRID 2011 and iLIDS-VID multi-shot re-identification datasets and demonstrate superior performance in comparison with the current state of the art.

Multi-scale Deep Learning Architectures for Person Re-identification

“…Since the primary focus of this paper is on image-based person re-id, we employ the simplest feature fusion scheme for video re-id: Given a video sequence, we compute features of each frame which are aggregated by max-pooling to form video level representation. In contrast, most of the state-ofthe-art video-based re-id methods [40,37,52,23,30,22] utilized the RNN models such as LSTM to perform temporal/sequence video feature fusion from each frame. Experimental settings.…”

Kernel cross-view collaborative representation based classification for person re-identification

“…Some works investigate the person re-identification problem using sparse or collaborative representations [22,31,16,24,25,26,47,53,23]. Lisanti et al [31] propose an Iterative Sparse Ranking (ISR) method that iteratively applies SRC with adaptive weighting strategies until ranking all the gallery images.…”

“…Differently from dictionary learning-based approaches [16,26], this work represents probe and gallery images using training samples. More importantly, different from previous works [31,22,16,26], we efficiently model the strong nonlinear transition of features between cameras achieving an analytical solution.…”

Robust Bayesian method for simultaneous block sparse signal recovery with applications to face recognition

“…In this work, we extend the SRC framework to the MMV case and consider performing FR when multiple images of the same subject, corrupted by non-stationary occlusions, are presented to the classifier. Our work is motivated, in part, by the person reidentification problem [12,13,14]. Srikrishna et al [14] addressed the re-identification problem by applying SSR to each individual image of the subject and aggregating the results to form a global classifier.…”

“…Srikrishna et al [14] addressed the re-identification problem by applying SSR to each individual image of the subject and aggregating the results to form a global classifier. As such, [14] did not address the MMV nature of the problem. The main motivation behind our work is to enforce the prior knowledge that the input images correspond to the same person within the SSR process, while still maintaining resilience to time-varying occlusions.…”