Deep Occlusion Reasoning for Multi-Camera Multi-Target Detection
lastmod 2026-11-10created 2019-03-122 edits
Deep Occlusion Reasoning for Multi-Camera Multi-Target Detection is a multi-object detection paper that deals with people occluding each other when the same scene is captured by multiple cameras. It extracts features from each viewpoint with a CNN and infers inter-camera occlusion relationships with a conditional random field (CRF) to estimate people's positions in crowded scenes. It belongs to the field of multi-camera pedestrian detection and tracking.
- Deep Occlusion Reasoning for Multi-Camera Multi-Target Detection
Documentation
- Deep Occlusion Reasoning for Multi-Camera Multi-Target Detection
- https://arxiv.org/abs/1704.05775