Active learning guided interactions for consistent image segmentation with reduced user interactions

Interactive techniques leverage the expert knowledge of users to produce accurate image segmentations. However, the segmentation accuracy varies with the users. Additionally, users may require training with the algorithm and its exposed parameters to obtain the best segmentation with minimal effort. Our work combines active learning with interactive segmentation and (i) achieves as good or better accuracy as a fully user guided segmentation with significantly lower number of user interactions (on average 50%), and (ii) achieves robust segmentation despite user variability. Our approach interacts with user to suggest placement of gestures. We present extensive experimental evaluation of our results on two different publicly available datasets.