Repulsion Loss: Detecting Pedestrians in a Crowd
arXiv:1711.07752
Abstract
Detecting individual pedestrians in a crowd remains a challenging problem since the pedestrians often gather together and occlude each other in real-world scenarios. In this paper, we first explore how a state-of-the-art pedestrian detector is harmed by crowd occlusion via experimentation, providing insights into the crowd occlusion problem. Then, we propose a novel bounding box regression loss specifically designed for crowd scenes, termed repulsion loss. This loss is driven by two motivations: the attraction by target, and the repulsion by other surrounding objects. The repulsion term prevents the proposal from shifting to surrounding objects thus leading to more crowd-robust localization. Our detector trained by repulsion loss outperforms all the state-of-the-art methods with a significant improvement in occlusion cases.
Accepted to IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2018
References in corpus (6)
Cited by in corpus (11)
- Object Detection in 20 Years: A Survey
- Acquisition of Localization Confidence for Accurate Object Detection
- Small-scale Pedestrian Detection Based on Somatic Topology Localization and Temporal Feature Aggregation
- Occlusion-aware R-CNN: Detecting Pedestrians in a Crowd
- Where, What, Whether: Multi-modal Learning Meets Pedestrian Detection
- A Survey on Deep Domain Adaptation and Tiny Object Detection Challenges, Techniques and Datasets
- Improving Object Detection with Inverted Attention
- PMC-GANs: Generating Multi-Scale High-Quality Pedestrian with Multimodal Cascaded GANs
- Pedestrian Detection with Autoregressive Network Phases
- DeepACEv2: Automated Chromosome Enumeration in Metaphase Cell Images Using Deep Convolutional Neural Networks
- Multi-channel Deep Supervision for Crowd Counting