Weakly Supervised Object Boundaries

TL;DR

This paper introduces a weakly supervised boundary detection method that uses only bounding box annotations to achieve high-quality object boundaries, reducing annotation costs and surpassing fully supervised methods.

Contribution

It presents a novel approach to generate boundary annotations from bounding boxes, enabling high-performance boundary detection without detailed boundary labels.

Findings

Achieves top performance on object boundary detection

Outperforms fully supervised state-of-the-art methods

Reduces annotation effort significantly

Abstract

State-of-the-art learning based boundary detection methods require extensive training data. Since labelling object boundaries is one of the most expensive types of annotations, there is a need to relax the requirement to carefully annotate images to make both the training more affordable and to extend the amount of training data. In this paper we propose a technique to generate weakly supervised annotations and show that bounding box annotations alone suffice to reach high-quality object boundaries without using any object-specific boundary annotations. With the proposed weak supervision techniques we achieve the top performance on the object boundary detection task, outperforming by a large margin the current fully supervised state-of-the-art methods.