Abstract
State-of-the-art learning based boundary detection methods require extensive training data.Since labelling ob-ject boundaries is one ofthe most expensive types of annota-tions,there is a need to relar 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 annota-tions 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 perfonn-ance on the object boundary detection task,outperfoming by a large margin the current fully supervised state-of-the-art methods.