Hierarchical Gaussian Descriptor for Person Re-identification

By Tetsu Matsukawa, Takahiro Okabe, Einoshin Suzuki, Yoichi Sato. Computer Vision and Pattern Recognition, 2016

579 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/0b3c67d90fc4a29d544ac2cf211e7732b04a2aa0

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

1. Yi Yang
2. Liang Zheng
3. Xiaogang Wang
4. Luc Van Gool
5. Dacheng Tao
6. Qi Tian
7. Ming-Hsuan Yang
8. Ling Shao

Citations in the corpus

• Deep Learning for Person Re-Identification: A Survey and Outlook (2020, background): “…[14], unconstrained poses [15], [16], [17], occlusions [18], [19], heterogeneous modalities [9], [20], etc. Early research efforts mainly focus on the hand-crafted feature construction with body structures [4], [21], [22], [23], [24] or distance metric learning [25], [26], [27], [28], [29], [30].”
• Deep Transfer Learning for Person Re-Identification (2016, background): “Earlier works focus on either designing viewinsensitive feature representations [2], [3], or learning an effective distance metric [4], [5], or both [6].”
• Orientation Invariant Feature Embedding and Spatial Temporal Regularization for Vehicle Re-identification (2017, background): “Many hand-crafted features are proposed to capture visual features for pedestrians [1,5,6,12,16,28].”
• A Siamese Long Short-Term Memory Architecture for Human Re-identification (2016, methodology): “For VIPeR dataset, several recent approaches [7,8,37,61,62] outperform our approach.”
• Image-Image Domain Adaptation with Preserved Self-Similarity and Domain-Dissimilarity for Person Re-identification (2017, background): “Hand-craft features [32, 14, 7, 33, 26, 51] can be directly employed for unsupervised”
• Beyond Triplet Loss: A Deep Quadruplet Network for Person Re-identification (2017, methodology): “Since VIPeR is relatively small, it is expected that deep learning might not be demonstrated to reach its full potential; instead, a hand-crafted metric learning may be more advantageous on this set, like GOG [23] and Ensembles [24].”
• Dynamic Label Graph Matching for Unsupervised Video Re-identification (2017, methodology): “In previous works, part from directly using hand-crafted descriptors [30, 14, 19, 16], some other unsupervised re-ID methods focus on finding shared invariant information (saliency [36] or dictionary [9, 22]) among cameras.”
• HydraPlus-Net: Attentive Deep Features for Pedestrian Analysis (2017, methodology): “As for the VIPeR [8] dataset, the null space semi-supervised learning method NFST [36], the similarity learning method SCSP [3], the hierarchical Gaussian model GOG+XQDA [19], and the triplet loss model TCP [5] are selected for comparison.”
• 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]”
• Attention-Aware Compositional Network for Person Re-identification (2018, background): “The first category mainly includes the traditional feature descriptors [53, 52, 51, 27, 7, 31, 35] and deep learning features [43, 41, 42, 12, 25, 40].”
• Person Re-identification by Deep Learning Multi-scale Representations (2017, background): “Many different handcrafted person image feature descriptors [13, 74, 61, 35, 70, 38, 29] have been designed in the past decade.”
• End-to-End Comparative Attention Networks for Person Re-Identification (2016, methodology): “We compare our model with the following state-of-the-art methods: SDALF [4], LMNN [5], ITML [1], KRMCA [45], LDM [3] , eSDC [10], Metric Ensembles (Ensembles) [8], KISSME [7], JointRe-id [16], FPNN [13], PersonNet [17], LOMO+XQDA [24], FT-JSTL+DGD [26], TCP [27], Embedding DM [35], LSSCDL [30] GOG [”
• Person Re-identification: Past, Present and Future (2016, background): “In [46], a hierarchical Gaussian feature is proposed to describe color and texture cues, which models each region by multiple Gaussian distributions.”
• Pose-Invariant Embedding for Deep Person Re-Identification (2017, methodology): “We first compare PIE (using Euclidean distance) with the latest unsupervised methods, e.g ., the Gaussian of Gaussian (GoG) [25], the Bag-of-Words (BOW) [42] descriptors, etc .”
• Enhancing Person Re-identification in a Self-Trained Subspace (2017, methodology): “We adopt the data splits in [3] for VIPeR, CUHK01, PRID450S, and GRID.”