Bounding Box Regression with Uncertainty for Accurate Object Detection

CVPR 2019 Yihui HeChenchen ZhuJianren WangMarios SavvidesXiangyu Zhang

Large-scale object detection datasets (e.g., MS-COCO) try to define the ground truth bounding boxes as clear as possible. However, we observe that ambiguities are still introduced when labeling the bounding boxes... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Object Detection COCO test-dev ResNet-50-FPN Mask R-CNN + KL Loss + var voting + soft-NMS box AP 40.4 # 53
Object Detection PASCAL VOC 2007 VGG-16 + KL Loss + var voting + soft-NMS MAP 71.6% # 22

Methods used in the Paper