Query-Guided End-To-End Person Search

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

doi:10.1109/cvpr.2019.00090

Cited by 124 publications

(115 citation statements)

References 47 publications

(137 reference statements)

Supporting

Mentioning

115

Contrasting

Order By: Relevance

“…Meng et al [9] learn the identity labels annotated at the untrimmed video level. Munjal et al [6] fused the query information into a Siamese network as a guide for global context information, proposal generation and similarity calculation. Another research trend is domain adaptation [44], [45] for PREID [5], [46], [47], where networks trained on the source domain can have a significant performance drop in the evaluation on the target domain [5].…”

Section: A Person Re-identificationmentioning

confidence: 99%

Multi Scale-Adaptive Super-Resolution Person Re-Identification Using GAN

et al. 2020

View full text Add to dashboard Cite

In real-world surveillance systems, the person images captured by the camera network consists of various low-resolution (LR) images. It creates a resolution mismatching problem when compared against high-resolution images of a targeted person. It significantly affects the performance of person re-Identification. This problem is known as Low-Resolution Person re-identification (LR PREID). An efficient strategy would be to exploit image super-resolution (SR) with person re-identification as a mutual learning approach. In this paper, we propose a novel method MSA-SR-PREID to solve this problem. The model takes low-resolution images on different resolutions and resized them to pre-defined fixed resolution. The design of the super-resolution network consists of ESRGAN and the de-Noising module to generate superresolution images. The SR images are later passed to the re-identification network to learn the unique descriptors to recognize a person identity. The performance of this model is evaluated on four competitive benchmarks, MLR-VIPeR, MLR-DukeMTMC-reID, VR-MSMT17, and VR-Market1501. The comparison with similar state-of-the-art demonstrates the superiority of our model.

show abstract

Section: A Person Re-identificationmentioning

confidence: 99%

Multi Scale-Adaptive Super-Resolution Person Re-Identification Using GAN

et al. 2020

View full text Add to dashboard Cite

show abstract

“…[21,22]. It is shown in [23,24] that one can utilize a joint formulation of reidentification and detection on full images, and still be on par with the state of the art. However, few later works [7,31] show that such a joint formulation degrades the detector performance, which is critical for a good tracking performance.…”

Section: Re-identificationmentioning

confidence: 99%

Joint detection and tracking in videos with identification features

Munjal¹,

Aftab²,

Amin³

et al. 2020

Image and Vision Computing

Self Cite

View full text Add to dashboard Cite

Recent works have shown that combining object detection and tracking tasks, in the case of video data, results in higher performance for both tasks, but they require a high frame-rate as a strict requirement for performance. This assumption is often violated in real-world applications, when models run on embedded devices, often at only a few frames per second. Videos at low frame-rate suffer from large object displacements. Here re-identification features may support to match large-displaced object detections, but current joint detection and re-identification formulations degrade the detector performance, as these two are contrasting tasks. In the real-world application having separate detector and re-id models is often not feasible, as both the memory and runtime effectively double. Towards robust long-term tracking applicable to reduced-computational-power devices, we propose the first joint optimization of detection, tracking and re-identification features for videos. Notably, our joint optimization maintains the detector performance, a typical multi-task challenge. At inference time, we leverage detections for tracking (tracking-by-detection) when the objects are visible, detectable and slowly moving in the image. We leverage instead re-identification features to match objects which disappeared (e.g. due to occlusion) for several frames or were not tracked due to fast motion (or low-frame-rate videos). Our proposed method reaches the state-of-the-art on MOT, it ranks 1st in the UA-DETRAC'18 tracking challenge among online trackers, and 3rd overall.

show abstract

“…Person Search. Person search is a recently developed task [2,3,36,50], which aims to match a target pedestrian from a great number of whole scene images. In the literature, there are two approaches to deal with this task.…”

Section: Related Workmentioning

confidence: 99%

“…Xiao et al [46] proposed an Individual Aggregation Network (IAN) that can accurately localize pedestrians by learning to minimize intra-person feature variations with a center loss. Other methods [2,28,36,50] focus on designing query-guided models for person search in an end-to-end manner. Munjal et al [36] proposed a query-guided end-to-end person search network (QEEPS), which utilized the query for its global context and local cues.…”

Section: Related Workmentioning

confidence: 99%

“…These two kinds of contexts provide different types of complementary information, which can be combined together to jointly help searching pedestrians. A few of works [2,28,36] make a preliminary effort on learning representation with contextual information, which simply utilize the contexts mixed with various information in the whole scene images, even some contexts are noise information (e.g., the other neighboring pedestrians only appear once in the scene images). The method [50] further employs neighboring co-travelers in the scene images as useful contextual information to improve the performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Dual Context-Aware Refinement Network for Person Search

Liu

Zha

Hong

et al. 2020

Proceedings of the 28th ACM International Conference on Multimedia

View full text Add to dashboard Cite

Person search has recently gained increasing attention as the novel task of localizing and identifying a target pedestrian from a gallery of non-cropped scene images. Its performance depends on accurate person detection and re-identification simultaneously by learning effective representations. In this work, we propose a novel dual context-aware refinement network (DCRNet) for person search, which jointly explores two kinds of contexts including intra-instance context and inter-instance context to learn discriminative representation. Specifically, an intra-instance context module is designed to refine the representation for the bounding box of a pedestrian by leveraging its surrounding regions covering the same pedestrian and its accessories, which contain abundant complementary visual appearance of pedestrians. Moreover, an inter-instance context module is proposed to expand the instance-level feature for the bounding box of a pedestrian, by utilizing the rich scene contexts of neighboring co-travelers across images. These two modules are built on top of a joint detection and feature learning framework, i.e., Faster R-CNN. Extensive experimental results on two challenging datasets have demonstrated the effectiveness of DCRNet with significant performance improvements over state-of-the-art methods. CCS CONCEPTS • Information systems → Top-k retrieval in databases.

show abstract

Query-Guided End-To-End Person Search

Cited by 124 publications

References 47 publications

Multi Scale-Adaptive Super-Resolution Person Re-Identification Using GAN

Multi Scale-Adaptive Super-Resolution Person Re-Identification Using GAN

Joint detection and tracking in videos with identification features

Dual Context-Aware Refinement Network for Person Search

Contact Info

Product

Resources

About