Learning Deep Feature Representations with Domain Guided Dropout for Person Re-identification

By Tong Xiao, Hongsheng Li, Wanli Ouyang, Xiaogang Wang. 2016

996 citations recorded by the source corpus.

Learning generic and robust feature representations with data from multiple domains for the same problem is of great value, especially for the problems that have multiple datasets but none of them are large enough to provide abundant data variations. In this work, we present a pipeline for learning deep feature representations from multiple domains with Convolutional Neural Networks (CNNs). When training a CNN with data from all the domains, some neurons learn representations shared across several domains, while some others are effective only for a specific one. Based on this important observation, we propose a Domain Guided Dropout algorithm to improve the feature learning procedure. Experiments show the effectiveness of our pipeline and the proposed algorithm. Our methods on the person re-identification problem outperform stateof-the-art methods on multiple datasets by large margins.

Subfields: 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 Computer Vision and Pattern Recognition

Citations in the corpus

Unsupervised Embedding Learning via Invariant and Spreading Instance Feature (2019, background): “Deep embedding learning usually learns an embedding function by minimizing the intra-class variation and maximizing the inter-class variation [32, 37, 45, 47].”
In Defense of the Triplet Loss for Person Re-Identification (2017, background): “However, using pretrained networks also leads to a design lock-in, and does not allow for the exploration of new deep learning advances or different architectures.”
Person Re-Identification by Deep Joint Learning of Multi-Loss Classification (2017, methodology): “Also, the DGD [Xiao et al. , 2016] model uses the classiﬁcation loss for model optimisation.”
Orientation Invariant Feature Embedding and Spatial Temporal Regularization for Vehicle Re-identification (2017, positive): “Recently, CNN-based features [2,21,22,27] have also achieved great progress on person ReID.”
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].”
Bilinear Attention Networks for Person Retrieval (2019, positive): “trained a model across multiple datasets [48] and used domain guided dropouts to mute domain-irrelevant neurons.”
Towards Backward-Compatible Representation Learning (2020, background): “Deep neural networks (DNNs) are widely applied to 1) learn embedding models using closed-world classification as a surrogate task [13, 29], using various forms of loss functions [35, 34, 39] and supervision methods [10, 13] to im-”
Improving Person Re-identification by Attribute and Identity Learning (2017, background): “Examples include [41, 49, 42, 9, 53].”
Deep Visual Domain Adaptation: A Survey (2018, background): “The techniques that are proven to be cost effective include adaptive batch normalization (BN) [58, 46, 57], weak-related weight [78], domain-guided dropout [102], and so forth.”
Scalable Person Re-identification on Supervised Smoothed Manifold (2017, neutral): “7 DGD [52] 75.”
Learning Deep Context-Aware Features over Body and Latent Parts for Person Re-identification (2017, methodology): “8%), and achieves comparable results with DGD [39] (38.”
Spindle Net: Person Re-identification with Human Body Region Guided Feature Decomposition and Fusion (2017, methodology): “The proposed Spindle Net is compared with several state-of-the-art ReID approaches in recent two years, including the metric learning method WARCA-χ2 [11], the null space semi-supervised learning method NFST [32], the deep neural network PersonNet [28], the Siamese CNN with learnable gate S-CNN [25]”
Distilled Person Re-Identification: Towards a More Scalable System (2019, background): “Person Reidentification has witnessed a fast growing development in recent years, from feature design [19, 15, 32, 31, 35, 69, 58] to distance metric learning [54, 19, 43, 25, 67, 36, 41, 30, 58, 33, 38, 31, 9, 68, 60, 63, 69, 29, 52, 6] and end-to-end deep learning [28, 1, 55, 57, 49, 22, 64, 65, 7”
Multi-level Factorisation Net for Person Re-identification (2018, methodology): “Deep Neural Networks for Person Re-ID Most recent person Re-ID methods train deep DNN models with various learning objectives including classification, verification and triplet ranking losses [42, 38, 24, 32, 27].”
Cross-View Asymmetric Metric Learning for Unsupervised Person Re-Identification (2017, methodology): “Data representation: In our experiments we used the deeplearning-based JSTL feature proposed in [32].”