High-level Semantic Feature Detection: A New Perspective for Pedestrian Detection

CVPR 2019 Wei LiuShengcai LiaoWeiqiang RenWeidong HuYinan Yu

Object detection generally requires sliding-window classifiers in tradition or anchor-based predictions in modern deep learning approaches. However, either of these approaches requires tedious configurations in windows or anchors... (read more)

PDF Abstract

Evaluation results from the paper


 SOTA for Pedestrian Detection on Caltech (using extra training data)

     Get a GitHub badge
Task Dataset Model Metric name Metric value Global rank Uses extra
training data
Compare
Pedestrian Detection Caltech CSP + CityPersons dataset Reasonable Miss Rate 3.8 # 2
Pedestrian Detection Caltech CSP Reasonable Miss Rate 4.5 # 5
Pedestrian Detection CityPersons CSP (with offset) + ResNet-50 Reasonable MR^-2 11.0 # 2
Pedestrian Detection CityPersons CSP (with offset) + ResNet-50 Heavy MR^-2 49.3 # 1
Pedestrian Detection CityPersons CSP (with offset) + ResNet-50 Partial MR^-2 10.4 # 1
Pedestrian Detection CityPersons CSP (with offset) + ResNet-50 Bare MR^-2 7.3 # 2
Pedestrian Detection CityPersons CSP (with offset) + ResNet-50 Small MR^-2 16.0 # 1
Pedestrian Detection CityPersons CSP (with offset) + ResNet-50 Medium MR^-2 3.7 # 1
Pedestrian Detection CityPersons CSP (with offset) + ResNet-50 Large MR^-2 6.5 # 1
Pedestrian Detection CityPersons CSP (with offset) + ResNet-50 Test Time 0.33s/img # 2