Multi-Stream Region Proposal Network for Pedestrian Detection

Pedestrian detection aims to locate person instances with bounding boxes in images or videos. Although researches have achieved significant progress in pedestrian detection recently, there are still open questions regarding occlusion, variation and confusion. In this paper, we propose a multi-stream region proposal network to enhance the ability of pedestrian detection. First, multiple visible region guided networks are proposed to obtain the disparate features based on diverse visible region patterns of human body. Then, fusion network is introduced to integrate the feature maps from multiple visible region patterns. Finally, a specially tailored region proposal network is applied to generate the proposal regions using the fused region features. In addition, boosted forest is used to classify the proposal regions. The proposed method is evaluated on Caltech database and achieves comparable performance with the state-of-the-art methods.