Query-guided End-to-End Person Search

Munjal, Bharti; Amin, Sikandar; Tombari, Federico; Galasso, Fabio

doi:10.48550/arxiv.1905.01203

Cited by 1 publication

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[49] utilizes GCN to explore the impact of context persons on the person search task. Nevertheless, we still achieve 2% gain on both mAP and top-1 accuracy compared with [43], and 2% improvement on mAP compared with [49], all of which prove the effectiveness of our method.…”

Section: Overall Person Search Performance On Cuhk-sysusupporting

confidence: 58%

“…Liu et al [42] recursively shrink the attentive regions till the target person is retrieved. In [43], global context of query-gallery pairs are emphasized by establishing a query-guided region proposal network and similarity sub-network in a siamese structure. Yan et al further improve the person search performance by exploiting co-travelers as global context.…”

Section: Person Searchmentioning

confidence: 99%

“…QEEPS [43] and GCNPS [49]), as well as the methods solving pedestrian detection and person identification separately (DSIFT+Euclidean [50], DSIFT+KISS-ME [51], BoW [52]+Cosine similarity, LO-MO+XQDA, and MGTS [15]).…”

Section: Comparison With State-of-the-art Approachesmentioning

confidence: 99%

“…NPSM [42] designs a person search approach by recursively shrinking the search area. QEEPS [43] proposes a strong person search framework by learning query-guided global context. [49] utilizes GCN to explore the impact of context persons on the person search task.…”

Section: Overall Person Search Performance On Cuhk-sysumentioning

confidence: 99%

See 3 more Smart Citations

Segmentation Mask Guided End-to-End Person Search

Zheng¹,

Xiao²,

Huang³

et al. 2019

Preprint

View full text Add to dashboard Cite

Person search aims to search for a target person among multiple images recorded by multiple surveillance cameras, which faces various challenges from both pedestrian detection and person re-identification. Besides the large intra-class variations owing to various illumination conditions, occlusions and varying poses, background clutters in the detected pedestrian bounding boxes further deteriorate the extracted features for each person, making them less discriminative. To tackle these problems, we develop a novel approach which guides the network with segmentation masks so that discriminative features can be learned invariant to the background clutters. We demonstrate that joint optimization of pedestrian detection, person re-identification and pedestrian segmentation enables to produce more discriminative features for pedestrian, and consequently leads to better person search performance. Extensive experiments on benchmark dataset CUHK-SYSU, show that our proposed model achieves the state-of-the-art performance with 86.3% mAP and 86.5% top-1 accuracy respectively.

show abstract

Section: Overall Person Search Performance On Cuhk-sysusupporting

confidence: 58%

Section: Person Searchmentioning

confidence: 99%