Deep-Person: Learning discriminative deep features for person Re-Identification

Bai, Xiang; Yang, Mingkun; Huang, Tong; Dou, Zhiyong; Yu, Rui; Xu, Yongchao

doi:10.1016/j.patcog.2019.107036

Cited by 206 publications

(125 citation statements)

References 47 publications

Supporting

Mentioning

125

Contrasting

Order By: Relevance

“…Local and global based approaches. Many approaches make use of both the global and local feature to simultaneously exploit their advantages [3,39,18,42,33,49,48]. Global features learned from the full image intend to capture the most discriminative clues of appearance but may fail to capture discriminative local details.…”

Section: Related Workmentioning

confidence: 99%

“…Multi-branch sub-networks (MB-Ns). Both global information and local details are important and complementary for re-ID [3,39,18,42,33,49]. In order to learn local detailed features of the separate region part rather than mixing all parts together, we adopt multi-branch sub-networks (MB-Ns) to learn local feature maps D i ∈ R h×w×c of size h × w with c channels, i = 1, 2, · · · , N , for N merged body part regions, respectively.…”

Section: Dsag-streammentioning

confidence: 99%

“…Network settings. We take ResNet-50 [11] to build our baseline networks as in some re-ID systems [3,35,48,2]. Similar to [35], the last spatial down-sample operation in the Conv5 layer is removed.…”

Section: Implementation Detailsmentioning

confidence: 99%

“…In recent years, many efforts have been made to alleviate these problems [33,53,37,48,34,19]. To make the features focus on some local details, some works make a straightforward partition of the person image into a few fixed rigid parts (e.g., horizontal stripes) and learn detailed local features [5,37,18,3,35,39]. However, such a partition cannot well align the human body parts.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Densely Semantically Aligned Person Re-Identification

Zhang

Lan

Zeng

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

276

189

View full text Add to dashboard Cite

We propose a densely semantically aligned person reidentification framework. It fundamentally addresses the body misalignment problem caused by pose/viewpoint variations, imperfect person detection, occlusion, etc. By leveraging the estimation of the dense semantics of a person image, we construct a set of densely semantically aligned part images (DSAP-images), where the same spatial positions have the same semantics across different images. We design a two-stream network that consists of a main full image stream (MF-Stream) and a densely semantically-aligned guiding stream (DSAG-Stream). The DSAG-Stream, with the DSAP-images as input, acts as a regulator to guide the MF-Stream to learn densely semantically aligned features from the original image. In the inference, the DSAG-Stream is discarded and only the MF-Stream is needed, which makes the inference system computationally efficient and robust. To the best of our knowledge, we are the first to make use of fine grained semantics to address the misalignment problems for re-ID. Our method achieves rank-1 accuracy of 78.9% (new protocol) on the CUHK03 dataset, 90.4% on the CUHK01 dataset, and 95.7% on the Mar-ket1501 dataset, outperforming state-of-the-art methods.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Dsag-streammentioning

confidence: 99%

Section: Implementation Detailsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Densely Semantically Aligned Person Re-Identification

Zhang

Lan

Zeng

et al. 2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

276

189

View full text Add to dashboard Cite

show abstract

“…In this section, we evaluate the proposed approach for the task of pedestrian re-identification on three benchmark datasets. Extensive comparisons are conducted and compared with methods include many state-of-the-art methods, such as GLAD [1], MSCAN [8], DLPA [5], SVDNet [12], PDC [13], TriNet [14], JLML [15], DML [10], DPFL [16], HA-CNN [17], GP-reid [18], PCB [6], Deep-Person [19], PCB+RPP [6], and AlignedReID [9].…”

Section: Methodsmentioning

confidence: 99%

Pedestrian re-Identification Based on Tree Branch Network with Local and Global Learning

Yang

Lai

et al. 2019

2019 IEEE International Conference on Multimedia and Expo (ICME)

View full text Add to dashboard Cite

Deep part-based methods in recent literature have revealed the great potential of learning local part-level representation for pedestrian image in the task of person re-identification. However, global features that capture discriminative holistic information of human body are usually ignored or not well exploited. This motivates us to investigate joint learning global and local features from pedestrian images. Specifically, in this work, we propose a novel framework termed tree branch network (TBN) for person re-identification. Given a pedestrain image, the feature maps generated by the backbone CNN, are partitioned recursively into serveral pieces, each of which is followed by a bottleneck structure that learns finer-grained features for each level in the hierarchical tree-like framework. In this way, represenations are learned in a coarseto-fine manner and finally assembled to produce more discriminative image descriptions. Experimental results demonstrate the effectiveness of the global and local feature learning method in the proposed TBN framework. We also show significant improvement in performance over state-of-the-art methods on three public benchmarks: Market-1501, CUHK-03 and DukeMTMC.

show abstract

Improving Deep Visual Representation for Person Re-identification by Global and Local Image-language Association

Chen

Liu

et al. 2018

Lecture Notes in Computer Science

129

View full text Add to dashboard Cite

Person re-identification is an important task that requires learning discriminative visual features for distinguishing different person identities. Diverse auxiliary information has been utilized to improve the visual feature learning. In this paper, we propose to exploit natural language description as additional training supervisions for effective visual features. Compared with other auxiliary information, language can describe a specific person from more compact and semantic visual aspects, thus is complementary to the pixel-level image data. Our method not only learns better global visual feature with the supervision of the overall description but also enforces semantic consistencies between local visual and linguistic features, which is achieved by building global and local image-language associations. The global image-language association is established according to the identity labels, while the local association is based upon the implicit correspondences between image regions and noun phrases. Extensive experiments demonstrate the effectiveness of employing language as training supervisions with the two association schemes. Our method achieves state-of-the-art performance without utilizing any auxiliary information during testing and shows better performance than other joint embedding methods for the image-language association.

show abstract

Deep-Person: Learning discriminative deep features for person Re-Identification

Cited by 206 publications

References 47 publications

Densely Semantically Aligned Person Re-Identification

Densely Semantically Aligned Person Re-Identification

Pedestrian re-Identification Based on Tree Branch Network with Local and Global Learning

Improving Deep Visual Representation for Person Re-identification by Global and Local Image-language Association

Contact Info

Product

Resources

About