Timezone: »

$\alpha$-IoU: A Family of Power Intersection over Union Losses for Bounding Box Regression
JIABO HE · Sarah Erfani · Xingjun Ma · James Bailey · Ying Chi · Xian-Sheng Hua

Fri Dec 10 08:30 AM -- 10:00 AM (PST) @
Bounding box (bbox) regression is a fundamental task in computer vision. So far, the most commonly used loss functions for bbox regression are the Intersection over Union (IoU) loss and its variants. In this paper, we generalize existing IoU-based losses to a new family of power IoU losses that have a power IoU term and an additional power regularization term with a single power parameter $\alpha$. We call this new family of losses the $\alpha$-IoU losses and analyze properties such as order preservingness and loss/gradient reweighting. Experiments on multiple object detection benchmarks and models demonstrate that $\alpha$-IoU losses, 1) can surpass existing IoU-based losses by a noticeable performance margin; 2) offer detectors more flexibility in achieving different levels of bbox regression accuracy by modulating $\alpha$; and 3) are more robust to small datasets and noisy bboxes.

Author Information

JIABO HE (the University of Melbourne)
Sarah Erfani (University of Melbourne)
Xingjun Ma (Deakin University)
Ying Chi (Alibaba DAMO Academy)
Xian-Sheng Hua (Alibaba Cloud)

Researcher/Senior Director of Alibaba Cloud, in charge of visual computing team in Big Data Department of Alibaba Cloud. He is an IEEE Fellow, ACM Distinguished Scientist and a TR35 Young Innovator Award Recipient, for his outstanding contributions in image/video analysis and search area. Joined Alibaba in April of 2015. Before that, he worked for Microsoft Research, Microsoft Bing and Microsoft Research Asia for 14 years. More details please visit: https://www.linkedin.com/in/xshua

More from the Same Authors