Intersection over Union with smoothing for bounding box regression

TL;DR

This paper introduces a smoothed IoU-based loss function for bounding box regression that converges faster, is more precise, and robust to noise, outperforming existing IoU variants.

Contribution

The paper proposes a novel smoothing technique for IoU loss, improving convergence speed, smoothness, and accuracy in bounding box regression tasks.

Findings

Outperforms IoU, DIoU, CIoU, and SIoU in experiments

Provides faster convergence and higher precision

Demonstrates robustness to noise in ground truth boxes

Abstract

We focus on the construction of a loss function for the bounding box regression. The Intersection over Union (IoU) metric is improved to converge faster, to make the surface of the loss function smooth and continuous over the whole searched space, and to reach a more precise approximation of the labels. The main principle is adding a smoothing part to the original IoU, where the smoothing part is given by a linear space with values that increases from the ground truth bounding box to the border of the input image, and thus covers the whole spatial search space. We show the motivation and formalism behind this loss function and experimentally prove that it outperforms IoU, DIoU, CIoU, and SIoU by a large margin. We experimentally show that the proposed loss function is robust with respect to the noise in the dimension of ground truth bounding boxes. The reference implementation is available at gitlab.com/irafm-ai/smoothing-iou.