Deep Cosine Metric Learning for Person Re-identification

By Nicolai Wojke, Alex Bewley. 2018

402 citations recorded by the source corpus.

Metric learning aims to construct an embedding where two extracted features corresponding to the same identity are likely to be closer than features from different identities. This paper presents a method for learning such a feature space where the cosine similarity is effectively optimized through a simple re-parametrization of the conventional softmax classification regime. At test time, the final classification layer can be stripped from the network to facilitate nearest neighbor queries on unseen individuals using the cosine similarity metric. This approach presents a simple alternative to direct metric learning objectives such as siamese networks that have required sophisticated pair or triplet sampling strategies in the past. The method is evaluated on two large-scale pedestrian re-identification datasets where competitive results are achieved overall. In particular, we achieve better generalization on the test set compared to a network trained with triplet loss.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Incremental Few-Shot Instance Segmentation (2021, methodology): “Our higher performance might also be due to our use of a cosine-similarity classiﬁer, which has been shown to produce more meaningful embeddings than the binary cross-entropy loss employed by Siamese Mask R-CNN [37].”
Learning From Demonstration in the Wild (2018, methodology): “To validate the ViBe vision module, we manually label 43 trajectories from the dataset and then compare its performance against two baselines, a simple IOU [18] tracker and Deep SORT [13], [14], a state-of-the-art MOT pipeline.”
Deep Learning for Person Re-Identification: A Survey and Outlook (2020, background): “Several works have also investigated the softmax variants [149], such as the sphere loss in [150] and AM softmax in [89].”
Deep Affinity Network for Multiple Object Tracking (2018, neutral): “affinity computation, DCML [97] and SSIM [98].”
Why Do Better Loss Functions Lead to Less Transferable Features? (2020, background): “Similar losses have appeared in previous literature [54, 71, 76, 72, 73, 21, 38, 15], and variants have introduced explicit additive or multiplicative margins to this loss that we do not consider here [38, 72, 73, 21].”
Iterative Deep Graph Learning for Graph Neural Networks: Better and Robust Node Embeddings (2020, background): “Common options for metric learning include cosine similarity [41, 50], radial basis function (RBF) kernel [54, 32] and attention mechanisms [47, 22].”