Abstract:With the development of self-driving technology, pedestrian detection models are becoming more useful and important. Although object detection models based on Convolutional Neural Networks have achieved favorable detection results, due to the occlusion in crowded scenarios, dense pedestrian detection is still a challenging task compared with generic object detection. For a CNN-based object detector, an appropriate label assignment method is significant. In this paper, we propose an effective training sample allocation method called Occlusion-Prediction aware Label Assignment (OPLA) to improve the performance of the detector in crowded scenarios. For those Ground Truths (GTs) without severe occlusion, we directly assign prediction boxes with the best matching quality as the corresponding positive samples. In terms of GTs in heavily crowded areas, we fine-tune the matching quality based on visible ratio to enhance the training sample selection priority of severe occluded GTs. In addition, we reveal the deficiency of currently popular post-processing algorithms, Non-Maximum Suppression (NMS) and Soft-NMS for dense pedestrian detection, and propose Hierarchical-NMS (H-NMS) whose strategy diverges with different Intersection over Union (IoU) levels. Experimental results on Crowdhuman, the largest dense pedestrian detection dataset at present, show our approach optimizes the Average Precision (AP) and Miss Rate (MR) of Fully Convolutional One Stage Object Detection (FCOS) by 4.05% and 5.35% and outperforms many state-of-the-art detectors, respectively. Besides, our method does not bring any increase in calculation during inference.

Dense Sigma set representation for pedestrian detection

Boosted Sigma Set for Pedestrian Detection

A Novel Approach to Design the Fast Pedestrian Detection for Video Surveillance System

A new depth descriptor for pedestrian detection in RGB-D images

Pedestrian detection based on modified dynamic background using gaussian mixture models and hog-svm detection

Towards Accurate Dense Pedestrian Detection Via Occlusion-Prediction Aware Label Assignment and Hierarchical-Nms.

Effective Pedestrian Detection Using Center-symmetric Local Binary/Trinary Patterns

Fast Pedestrian Detection And Tracking Based On Vibe Combined Hog-Svm Scheme

Pedestrian Detection Inspired by Appearance Constancy and Shape Symmetry

Robust pedestrian detection in thermal infrared imagery using a shape distribution histogram feature and modified sparse representation classification

Pyramid Center-Symmetric Local Binary/Trinary Patterns for Effective Pedestrian Detection

Pedestrian Target Recognition Algorithm in Public Places Based on Representation Learning and Similarity Learning

S 3 D: Scalable Pedestrian Detection via Score Scale Surface Discrimination

Sigma Set: A Small Second Order Statistical Region Descriptor

Feature Selection And Pedestrian Detection Based On Sparse Representation

Pedestrian Detection in Complex Scene Using Full Binary Tree Classifiers Based on Locally Assembled Binary Haar-Like Features

Pedestrian Detection Using a Mixture Mask Model

An Experimental Evaluation of Local Features for Pedestrian Classification

Pyramid Center-symmetric Local 1 Binary / Trinary Patterns for Pedestrian 2 Detection

An Experimental Study on Pedestrian Classification Using Local Features.

Performance Evaluation Of Local Features In Human Classification And Detection