Beyond Triplet Loss: A Deep Quadruplet Network for Person Re-identification

By Weihua Chen, Xiaotang Chen, Jianguo Zhang, Kaiqi Huang. 2017

1,219 citations recorded by the source corpus.

Person re-identification (ReID) is an important task in wide area video surveillance which focuses on identifying people across different cameras. Recently, deep learning networks with a triplet loss become a common framework for person ReID. However, the triplet loss pays main attentions on obtaining correct orders on the training set. It still suffers from a weaker generalization capability from the training set to the testing set, thus resulting in inferior performance. In this paper, we design a quadruplet loss, which can lead to the model output with a larger inter-class variation and a smaller intra-class variation compared to the triplet loss. As a result, our model has a better generalization ability and can achieve a higher performance on the testing set. In particular, a quadruplet deep network using a margin-based online hard negative mining is proposed based on the quadruplet loss for the person ReID. In extensive experiments, the proposed network outperforms most of the state-of-the-art algorithms on representative datasets which clearly demonstrates the effectiveness of our proposed method.

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

Semantic Drift Compensation for Class-Incremental Learning (2020, methodology): “A direct extension to this method is a quadruplet network [7], where the number of points is extended to four and three pairs are used at once in a loss function.”
CosFace: Large Margin Cosine Loss for Deep Face Recognition (2018, background): “Other loss functions [47, 6, 4, 3] based on contrastive loss or center loss also demonstrate the performance on enhancing discrimination.”
Continual Representation Learning for Biometric Identification (2020, background): “Biometric identification [23,51], including face recognition [11, 37, 67] and person re-identification (re-id) [34, 66, 71], has achieved significant progress in the recent years due to the advances in modern learnable representations [8, 9, 11, 20, 34, 52, 60, 67] and emerging large datasets [17, 2”
Fine-Grained Action Retrieval Through Multiple Parts-of-Speech Embeddings (2019, background): “Recently, neural networks trained with a ranking loss considering image pairs [27], triplets [35], quadruplets [5] or beyond [32], have been considered for metric learning [17, 35] and for a broad range of search tasks such as face/person identiﬁcation [29, 5, 16, 2] or instance retrieval [10, 27].”
Self-Supervised Learning on Graphs: Contrastive, Generative, or Predictive (2021, background): “The quadruplet loss [56] further considers imposing constraints on inter-class samples on top of the triplet margin loss, defined as:”
Mancs: A Multi-task Attentional Network with Curriculum Sampling for Person Re-Identification (2018, methodology): “Except triplet loss, contrastive loss [41] and quadruplet loss [6] are also used in person re-ID task.”
Attention-Aware Compositional Network for Person Re-identification (2018, methodology): “One set is poseirrelevant, which includes the null space semi-supervised learning method (NFST) [48], the domain guided dropout method (JSTL) [43], the deep transfer learning method (Transfer) [16], the Singular Vector Decomposition method (SVDNet) [38], the Online Instance Matching (OIM) method [44”
Person Recognition in Personal Photo Collections (2017, background): “There is an active line of research on pedestrian re-id, starting with hand-crafted features [30], [31], [32] and has moved towards deep feature based schemes [28], [33], [34], [35], [36], [37], [38], [39], [40].”
Memory-Based Neighbourhood Embedding for Visual Recognition (2019, methodology): “At the same time, there also exists a stream of methods which supervise the feature embeddings by imposing different loss functions, including the contrastive loss [8], the triplet loss [45] and quadruplet loss [3].”
Mask-Guided Contrastive Attention Model for Person Re-identification (2018, background): “In addition, some works try to introduce the pair-wise contrastive loss [14], triplet ranking loss [54] and quadruplet loss [8] to further enhance the IDE feature.”
Unsupervised Person Re-identification (2017, methodology): “These methods directly output the similarity score of two input images by focusing on their distinct body patches, without an explicit feature extraction process.”
A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunities (2022, background): “E-nagivate sample [173] CVPR’17 Hard Sample Fully-Connected Layers Not only considers the relative distance between positive and negative samples, but also adds the absolute distance between positive samples”
Characterizing Generalization under Out-Of-Distribution Shifts in Deep Metric Learning (2021, background): “Proposed approaches most commonly rely on a surrogate ranking task over tuples during training [62], ranging from simple pairs [17] and triplets [57] to higher-order quadruplets [5] and more generic n-tuples [61, 43, 22, 70].”
Hyperbolic Image Embeddings (2019, methodology): “Some methods use the combination of the aforementioned approaches: [4] adopts veriﬁcation losses that are applied to the similarities predicted by the special sub-network instead of the simple Euclidean similarities.”
Dual Attention Matching Network for Context-Aware Feature Sequence Based Person Re-identification (2018, methodology): “…feature vectors; in [28, 39, 49], local features of recurrent appearance data are extracted and integrated using temporal-pooling strategy; in [14, 5], to enhance the generalization capability of the learned embed-dings, the pairwise similarity criterion is extended to triplet or quadruplet.”
Camera Style Adaptation for Person Re-identification (2017, neutral): “Many deep learning methods [38, 34, 33, 3, 24] have been proposed in person re-ID.”
AANet: Attribute Attention Network for Person Re-Identifications (2019, neutral): “Various approaches such as triplet and quadruplet losses have been employed to learn such embedding spaces [10, 2].”