CircleNet: Reciprocating Feature Adaptation for Robust Pedestrian Detection

TL;DR

CircleNet introduces a novel feature adaptation model for pedestrian detection that mimics human focus on occluded and low-resolution objects through iterative, reciprocal feature refinement, significantly improving detection robustness.

Contribution

The paper presents CircleNet, a new feature learning framework with reciprocal feature adaptation and an instance decomposition training strategy for robust pedestrian detection.

Findings

Improves detection of occluded pedestrians

Enhances low-resolution pedestrian detection

Maintains performance on normal instances

Abstract

Pedestrian detection in the wild remains a challenging problem especially when the scene contains significant occlusion and/or low resolution of the pedestrians to be detected. Existing methods are unable to adapt to these difficult cases while maintaining acceptable performance. In this paper we propose a novel feature learning model, referred to as CircleNet, to achieve feature adaptation by mimicking the process humans looking at low resolution and occluded objects: focusing on it again, at a finer scale, if the object can not be identified clearly for the first time. CircleNet is implemented as a set of feature pyramids and uses weight sharing path augmentation for better feature fusion. It targets at reciprocating feature adaptation and iterative object detection using multiple top-down and bottom-up pathways. To take full advantage of the feature adaptation capability in CircleNet, we design an instance decomposition training strategy to focus on detecting pedestrian instances of various resolutions and different occlusion levels in each cycle. Specifically, CircleNet implements feature ensemble with the idea of hard negative boosting in an end-to-end manner. Experiments on two pedestrian detection datasets, Caltech and CityPersons, show that CircleNet improves the performance of occluded and low-resolution pedestrians with significant margins while maintaining good performance on normal instances.