Abstract:The person re-identification（re-ID） community has witnessed an explosion in the scale of data that it has to handle. On one hand, it is important for large-scale re-ID to provide constant or sublinear search time and dramatically reduce the storage cost for data points from the viewpoint of efficiency. On the other hand, the semantic affinity existing in the original space should be preserved because it greatly boosts the accuracy of re-ID. To this end, we use the deep hashing method, which utilizes the pairwise similarity and classification label to learn deep hash mapping functions, in order to provide discriminative representations. More importantly, considering the great advantage of asymmetric hashing over the existing symmetric one, we finally propose an asymmetric deep hashing（ADH） method for large-scale re-ID. Specifically, a two-stream asymmetric convolutional neural network is constructed to learn the similarity between image pairs. Another asymmetric pairwise loss is formulated to capture the similarity between the binary hashing codes and real-value representations derived from the deep hash mapping functions,so as to constrain the binary hash codes in the Hamming space to preserve the semantic structure existing in the original space. Then, the image labels are further explored to have a direct impact on the hash function learning through a classification loss. Furthermore, an efficient alternating algorithm is elaborately designed to jointly optimize the asymmetric deep hash functions and high-quality binary codes, by optimizing one parameter with the other parameters fixed. Experiments on the four benchmarks, i.e., DukeMTMC-reID, Market-1501, Market-1501+500 k,and CUHK03 substantiate the competitive accuracy and superior efficiency of the proposed ADH over the compared state-of-the-art methods for large-scale re-ID.

Person Re-Identification By Distance Metric Learning To Discrete Hashing

Discrete Cross-Modal Hashing for Efficient Multimedia Retrieval

Deep Multi-Index Hashing for Person Re-Identification.

Hashing person re-ID with self-distilling smooth relaxation.

Asymmetric Deep Hashing for Person Re-Identifications

People Re-Identification with Local Distance Comparison Using Learned Metric.

Extending Hashing Towards Fast Re-Identification

Person Re-identification Based on Hash.

Adaptive Metric Learning For People Re-Identification

Adaptive Metric Learning in Local Distance Comparison for People Re-identification

Anchor-Supported Multi-Modality Hashing Embedding For Person Re-Identification

Bit-Scalable Deep Hashing with Regularized Similarity Learning for Image Retrieval and Person Re-Identification

Fast Person Re-identification Via Cross-Camera Semantic Binary Transformation

Salience-Guided Iterative Asymmetric Mutual Hashing for Fast Person Re-Identification

Person Re-Identification by Discriminant Analytical Least Squares Metric Learning

Person Re-Identification by Enhanced Local Maximal Occurrence Representation and Generalized Similarity Metric Learning

Discrete Metric Learning for Fast Image Set Classification

Self-Guided Hash Coding for Large-Scale Person Re-identification

Deep View-Aware Metric Learning for Person Re-Identification

Person Re-identification Using Data-Driven Metric Adaptation.

Relevance Metric Learning for Person Re-identification by Exploiting Global Similarities