Filters








80 Hits in 5.2 sec

Multiple Expert Brainstorming for Domain Adaptive Person Re-identification [article]

Yunpeng Zhai, Qixiang Ye, Shijian Lu, Mengxi Jia, Rongrong Ji, Yonghong Tian
2020 arXiv   pre-print
In this paper, we propose a multiple expert brainstorming network (MEB-Net) for domain adaptive person re-ID, opening up a promising direction about model ensemble problem under unsupervised conditions  ...  Often the best performing deep neural models are ensembles of multiple base-level networks, nevertheless, ensemble learning with respect to domain adaptive person re-ID remains unexplored.  ...  Introduction Person re-identification (re-ID) aims to match persons in an image gallery collected from non-overlapping camera networks [40] , [14] , [16] .  ... 
arXiv:2007.01546v3 fatcat:m66wzjbigjbxtohccg4psa4oue

Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification [article]

Yixiao Ge, Dapeng Chen, Hongsheng Li
2020 arXiv   pre-print
Person re-identification (re-ID) aims at identifying the same persons' images across different cameras.  ...  In addition, the common practice is to adopt both the classification loss and the triplet loss jointly for achieving optimal performances in person re-ID models.  ...  INTRODUCTION Person re-identification (re-ID) aims at retrieving the same persons' images from images captured by different cameras.  ... 
arXiv:2001.01526v2 fatcat:7nqggtsmwvhp3kkz5ofm76xuli

Unsupervised and self-adaptative techniques for cross-domain person re-identification [article]

Gabriel Bertocco and Fernanda Andaló and Anderson Rocha
2021 arXiv   pre-print
Person Re-Identification (ReID) across non-overlapping cameras is a challenging task and, for this reason, most works in the prior art rely on supervised feature learning from a labeled dataset to match  ...  We also introduce a new self-ensembling strategy, in which weights from different iterations are aggregated to create a final model combining knowledge from distinct moments of the adaptation.  ...  RELATED WORK Several works have been proposed to address Unsupervised Domain Adaptation for Person Re-Identification.  ... 
arXiv:2103.11520v2 fatcat:dm3ribn4fjftxg33kd2xeqggoy

Dual-Triplet Metric Learning for Unsupervised Domain Adaptation in Video-Based Face Recognition [article]

George Ekladious, Hugo Lemoine, Eric Granger, Kaveh Kamali, Salim Moudache
2020 arXiv   pre-print
In many video surveillance applications, like face recognition (FR) and person re-identification, a pair-wise matcher is used to assign a query image captured using a video camera to the corresponding  ...  Then, the student relies on the teacher to iteratively label the positive and negative pairs collected during, e.g., initial camera calibration.  ...  In many video surveillance applications, like face recognition (FR) and person re-identification, a pair-wise matcher is used to assign a query image captured using a video camera to the corresponding  ... 
arXiv:2002.04206v1 fatcat:oi5skqtdxjaf3bt6ht6hgpxpry

DarkRank: Accelerating Deep Metric Learning via Cross Sample Similarities Transfer [article]

Yuntao Chen, Naiyan Wang, Zhaoxiang Zhang
2017 arXiv   pre-print
We test our proposed DarkRank method on various metric learning tasks including pedestrian re-identification, image retrieval and image clustering. The results are quite encouraging.  ...  Recently, Hinton etal. have shown that the dark knowledge within a powerful teacher model can significantly help the training of a smaller and faster student network.  ...  Experiments In this section, we test the performance of our DarkRank method on several metric learning tasks including person re-identification, image retrieval and clustering, and compare it with several  ... 
arXiv:1707.01220v2 fatcat:3bwxube6djd3nhxtg3nfexxpom

Deep Learning for Person Re-identification: A Survey and Outlook [article]

Mang Ye, Jianbing Shen, Gaojie Lin, Tao Xiang, Ling Shao, Steven C. H. Hoi
2021 arXiv   pre-print
Person re-identification (Re-ID) aims at retrieving a person of interest across multiple non-overlapping cameras.  ...  Meanwhile, we introduce a new evaluation metric (mINP) for person Re-ID, indicating the cost for finding all the correct matches, which provides an additional criteria to evaluate the Re-ID system for  ...  Model distillation is another approach, e.g., a multi-teacher adaptive similarity distillation framework is proposed in [239] , which learns a user-specified lightweight student model from multiple teacher  ... 
arXiv:2001.04193v2 fatcat:4d3thmsr3va2tnu72nawlu2wxy

Robust, Extensible, and Fast: Teamed Classifiers for Vehicle Tracking and Vehicle Re-ID in Multi-Camera Networks [article]

Abhijit Suprem, Rodrigo Alves Lima, Bruno Padilha, Joao Eduardo Ferreira, Calton Pu
2020 arXiv   pre-print
We describe an implementation for vehicle tracking and vehicle re-identification (re-id), where we implement a zero-shot learning (ZSL) system that performs automated tracking of all vehicles all the time  ...  This includes performing tasks such as object detection, attribute identification, and vehicle/person tracking across different cameras without overlap.  ...  Since we adapt student teacher networks for the detector team and standard image classifiers for some attribute extractors, we focus evaluation on the novel team models: the brand discrimination team and  ... 
arXiv:1912.04423v2 fatcat:yjpl73aqezewhmkvvut4ziet3i

Self-Supervised Small Soccer Player Detection and Tracking [article]

Samuel Hurault, Coloma Ballester, Gloria Haro
2020 arXiv   pre-print
Although a straightforward solution would be to retrain these models by using a more specific dataset, the lack of such publicly available annotated datasets entails searching for other effective solutions  ...  State-of-the-art tracking algorithms achieve impressive results in scenarios on which they have been trained for, but they fail in challenging ones such as soccer games.  ...  For these reasons, we consider a complementary visual consistency measure. 3.2.2 Re-identification and visual consistency.  ... 
arXiv:2011.10336v1 fatcat:ntegz37byvcfxftecjx2nxwq4y

SSKD: Self-Supervised Knowledge Distillation for Cross Domain Adaptive Person Re-Identification [article]

Junhui Yin, Jiayan Qiu, Siqing Zhang, Zhanyu Ma, Jun Guo
2020 arXiv   pre-print
Domain adaptive person re-identification (re-ID) is a challenging task due to the large discrepancy between the source domain and the target domain.  ...  Finally, the two modules can resist label noise for re-ID by enhancing each other and systematically integrating label information from unlabeled images.  ...  Duke → Market Market → Duke mAP top-1 top-5 top-10 mAP top-1 top-5 top-10 Pseudo label refinery for unsupervised domain adaptation on person re-identification .  ... 
arXiv:2009.05972v1 fatcat:hubaa3jdw5eillbwsxtbj66cva

Low-Resolution Face Recognition in the Wild via Selective Knowledge Distillation

Shiming Ge, Shengwei Zhao, Chenyu Li, Jia Li
2019 IEEE Transactions on Image Processing  
The teacher stream is represented by a complex CNN for high-accuracy recognition, and the student stream is represented by a much simpler CNN for low-complexity recognition.  ...  To avoid significant performance drop at the student stream, we then selectively distil the most informative facial features from the teacher stream by solving a sparse graph optimization problem, which  ...  To verify that, we use the student model S-16-sc pre-trained on UMDFaces to fine-tune a new model for face identification on UCCS.  ... 
doi:10.1109/tip.2018.2883743 fatcat:32z2fr6vpzbn3esvl5uwqrgd6e

Person search: New paradigm of person re-identification: A survey and outlook of recent works

Khawar Islam
2020 Image and Vision Computing  
In last few years, deep learning has played unremarkable role for the solution of re-identification problem. Deep learning shows incredible performance in person (re-ID) and search.  ...  This task aims to find a probe person from whole scene which shows great significance in video surveillance field to track lost people, re-identification, and verification of person.  ...  Distilled QEEPS [64] Introduced new approach called knowledge distillation and proposed two separate methods called teacher and student frameworks, for the person detection and for re-ID parts.  ... 
doi:10.1016/j.imavis.2020.103970 fatcat:g2zuqww7tbdszkxrc2wkrfno2y

Improving Unsupervised Domain Adaptive Re-Identification via Source-Guided Selection of Pseudo-Labeling Hyperparameters [article]

Fabian Dubourvieux, Angélique Loesch, Romaric Audigier, Samia Ainouz, Stéphane Canu
2021 arXiv   pre-print
Unsupervised Domain Adaptation (UDA) for re-identification (re-ID) is a challenging task: to avoid a costly annotation of additional data, it aims at transferring knowledge from a domain with annotated  ...  Experiments on commonly used person re-ID and vehicle re-ID datasets show that our proposed HyPASS consistently improves the best state-of-the-art methods in re-ID compared to the commonly used empirical  ...  Most of these works improve the classical selflearning algorithm on not overfitting the pseudo-label errors, by using teacher-student or ensemble of expert models [13, 64, 61] while other approaches  ... 
arXiv:2110.07897v1 fatcat:cdkohbzlwjg5plmume6gw32ks4

Person Re-IDentification based on mutual learning with embedded noise block

Xinyue Fan, Jia Zhang, Yang Lin
2021 IEEE Access  
INDEX TERMS Person re-identification, mutual learning, knowledge distillation, network decoupling  ...  Some person re-identification(Re-ID) algorithms based on deep learning utilizes a baseline as basis to modify, and add some strategies to achieve better performance.  ...  unsupervised person Re-ID.  ... 
doi:10.1109/access.2021.3102450 fatcat:u2au3zd7tjeb5ituomstvm7sde

2020 Index IEEE Transactions on Image Processing Vol. 29

2020 IEEE Transactions on Image Processing  
., +, TIP 2020 5805-5816 Fine-Grained Spatial Alignment Model for Person Re-Identification With Focal Triplet Loss.  ...  Zhang, P., +, TIP 2020 29-43 Fine-Grained Spatial Alignment Model for Person Re-Identification With Focal Triplet Loss.  ... 
doi:10.1109/tip.2020.3046056 fatcat:24m6k2elprf2nfmucbjzhvzk3m

Bridging Gap between Image Pixels and Semantics via Supervision: A Survey [article]

Jiali Duan, C.-C. Jay Kuo
2021 arXiv   pre-print
Experiences are drawn from two application domains to illustrate this point: 1) object detection and 2) metric learning for content-based image retrieval (CBIR).  ...  The fact that there exists a gap between low-level features and semantic meanings of images, called the semantic gap, is known for decades. Resolution of the semantic gap is a long standing problem.  ...  A teacher-student semisupervised image retrieval method was presented in [197] , where the pairwise information learned by the teacher network is used as the guidance to train the student network.  ... 
arXiv:2107.13757v2 fatcat:zk4vatjzxzfxfoqy2ojwx4gv2i
« Previous Showing results 1 — 15 out of 80 results