Single-Pedestrian Detection Aided by Multi-pedestrian Detection

By Wanli Ouyang, Xiaogang Wang. 2013

134 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Learning Non-maximum Suppression (2017, background): “One line of work proposes to detect pairs of objects instead of each individual objects in order to handle strong occlusion [24, 29, 19].”
DeepID-Net: multi-stage and deformable deep convolutional neural networks for object detection (2014, methodology): “Deformable part-based models were used in [20, 65, 41, 39] for handling translational movement of parts.”
DeepID-Net: Object Detection with Deformable Part Based Convolutional Neural Networks (2017, background): “Part deformation handling is a key factor for the recent progress in object detection [12], [83], [13], [73], [37].”
How Far are We from Solving Pedestrian Detection? (2016, background): “It also shows that optical flow [18] and context information [16] are complementary to image features and can further boost detection accuracy.”
Pedestrian detection with a Large-Field-Of-View deep network (2015, methodology): “For example, Ouyang and Wang [31] use two-pedestrian detection to improve on single-pedestrian detection, Ouyang et al [41] jointly detect two neighbouring pedestrians.”
Real-Time Pedestrian Detection with Deep Network Cascades (2015, background): “These approaches are still slow, ranging from over a second per image [25] to several minutes [37].”
Detection and Tracking of Occluded People (2014, background): “2012), several recent publications proposed approaches for detection and tracking of occluded people (Pepik et al. 2013; Ouyang and Wang 2013; Tang et al. 2013).”
EuroCity Persons: A Novel Benchmark for Person Detection in Traffic Scenes (2018, methodology): “Deformable Part Models (DPM) using Histograms of Oriented Gradients (HOG) features [36], [37], [38], and Decision Forests using ICF features [39], [40], [41], [42] were until a few years ago the established pedestrian detection methods [12].”
Deep Learning Strong Parts for Pedestrian Detection (2015, methodology): “[24] employed a probabilistic framework to model the relationship between the configurations estimated by single- and multi-pedestrian detectors.”
Recent Advances in Deep Learning for Object Detection (2019, background): “[225] addressed this issue 1706 from another aspect.”
Ten Years of Pedestrian Detection, What Have We Learned? (2014, methodology): “The context information is injected using the +2Ped re-weighting strategy [35] (the detection scores are combined with the scores of a “2 person” DPM detector).”
Occlusion-aware R-CNN: Detecting Pedestrians in a Crowd (2018, background): “[37] exploit multi-pedestrian detectors to aid single-pedestrian detectors to handle partial occlusions, especially when the pedestrians gather together and occlude each other in real-world scenarios.”
A Convnet for Non-maximum Suppression (2015, methodology): “Co-occurrence To better handle dense crowds or common object pairs, it has been proposed to use specialized 2-object detectors (Sadeghi & Farhadi, 2011; Tang et al., 2012; Ouyang & Wang, 2013), which then require a more careful NMS strategy to merge single-object with double-object detections.”