Improving Object Detection with Selective Self-supervised Self-training

TL;DR

This paper enhances object detection by leveraging Web images through a selective self-supervised self-training approach, addressing domain gaps and improving detection accuracy on challenging object classes.

Contribution

It introduces a novel selective net to refine supervision signals from Web images, enabling effective use of unlabeled data for object detection.

Findings

Achieved state-of-the-art results on backpacks and chairs detection.

Web images with selective supervision improve detection accuracy.

The method effectively handles domain gaps between Web and curated data.

Abstract

We study how to leverage Web images to augment human-curated object detection datasets. Our approach is two-pronged. On the one hand, we retrieve Web images by image-to-image search, which incurs less domain shift from the curated data than other search methods. The Web images are diverse, supplying a wide variety of object poses, appearances, their interactions with the context, etc. On the other hand, we propose a novel learning method motivated by two parallel lines of work that explore unlabeled data for image classification: self-training and self-supervised learning. They fail to improve object detectors in their vanilla forms due to the domain gap between the Web images and curated datasets. To tackle this challenge, we propose a selective net to rectify the supervision signals in Web images. It not only identifies positive bounding boxes but also creates a safe zone for mining hard negative boxes. We report state-of-the-art results on detecting backpacks and chairs from everyday scenes, along with other challenging object classes.