Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regression

Code

• github.com/gau-nernst/CenterNet
  pytorch★ 71
• github.com/OFRIN/Tensorflow_GIoU
  tf★ 0
• github.com/RuiminChen/GIouloss_CIouloss_caffe
  none★ 0
• github.com/JaryHuang/awesome_SSD_FPN_GIoU
  pytorch★ 0
• github.com/LinRiver/YOLOv3-on-LISA-Traffic-Sign-Detection-with-darknet
  none★ 0
• github.com/sremes/a2d2
  tf★ 0
• github.com/RuiminChen/GIou_loss_caffe
  none★ 0
• github.com/kalubin-awym/GIoU-loss-for-RetinaNet
  none★ 0
• github.com/AnselmC/bamot
  none★ 0

Abstract

Intersection over Union (IoU) is the most popular evaluation metric used in the object detection benchmarks. However, there is a gap between optimizing the commonly used distance losses for regressing the parameters of a bounding box and maximizing this metric value. The optimal objective for a metric is the metric itself. In the case of axis-aligned 2D bounding boxes, it can be shown that IoU can be directly used as a regression loss. However, IoU has a plateau making it infeasible to optimize in the case of non-overlapping bounding boxes. In this paper, we address the weaknesses of IoU by introducing a generalized version as both a new loss and a new metric. By incorporating this generalized IoU (GIoU) as a loss into the state-of-the art object detection frameworks, we show a consistent improvement on their performance using both the standard, IoU based, and new, GIoU based, performance measures on popular object detection benchmarks such as PASCAL VOC and MS COCO.

Tasks

Reproductions