Learning Transformations To Reduce the Geometric Shift in Object Detection

TL;DR

This paper proposes a self-training method to learn geometric transformations that reduce performance drops in object detection caused by geometric shifts, without requiring labeled data or camera info.

Contribution

It introduces a novel self-training approach to learn geometric transformations that mitigate geometric domain shifts in object detection.

Findings

Improved detection performance under camera FoV changes

Enhanced robustness to viewpoint variations

Effective without labeled target data or camera details

Abstract

The performance of modern object detectors drops when the test distribution differs from the training one. Most of the methods that address this focus on object appearance changes caused by, e.g., different illumination conditions, or gaps between synthetic and real images. Here, by contrast, we tackle geometric shifts emerging from variations in the image capture process, or due to the constraints of the environment causing differences in the apparent geometry of the content itself. We introduce a self-training approach that learns a set of geometric transformations to minimize these shifts without leveraging any labeled data in the new domain, nor any information about the cameras. We evaluate our method on two different shifts, i.e., a camera's field of view (FoV) change and a viewpoint change. Our results evidence that learning geometric transformations helps detectors to perform better in the target domains.