PS-RCNN: Detecting Secondary Human Instances in a Crowd via Primary Object Suppression

TL;DR

This paper introduces PS-RCNN, a two-stage detector that improves the detection of heavily occluded humans in crowded scenes by combining primary object detection, suppression, and a specialized second detection stage, enhanced with high-resolution features.

Contribution

The paper presents a novel two-stage detection framework with human-shaped masks and high-resolution RoI features to better detect occluded humans in crowded scenes, outperforming baseline methods.

Findings

Improves recall and AP by 4.49% and 2.92% on CrowdHuman dataset.

Effective in detecting heavily occluded human instances.

Achieves similar improvements on Widerperson dataset.

Abstract

Detecting human bodies in highly crowded scenes is a challenging problem. Two main reasons result in such a problem: 1). weak visual cues of heavily occluded instances can hardly provide sufficient information for accurate detection; 2). heavily occluded instances are easier to be suppressed by Non-Maximum-Suppression (NMS). To address these two issues, we introduce a variant of two-stage detectors called PS-RCNN. PS-RCNN first detects slightly/none occluded objects by an R-CNN module (referred as P-RCNN), and then suppress the detected instances by human-shaped masks so that the features of heavily occluded instances can stand out. After that, PS-RCNN utilizes another R-CNN module specialized in heavily occluded human detection (referred as S-RCNN) to detect the rest missed objects by P-RCNN. Final results are the ensemble of the outputs from these two R-CNNs. Moreover, we introduce a High Resolution RoI Align (HRRA) module to retain as much of fine-grained features of visible parts of the heavily occluded humans as possible. Our PS-RCNN significantly improves recall and AP by 4.49% and 2.92% respectively on CrowdHuman, compared to the baseline. Similar improvements on Widerperson are also achieved by the PS-RCNN.