Seed, Expand and Constrain: Three Principles for Weakly-Supervised Image Segmentation
This addresses the problem of reducing annotation costs for image segmentation, though it is incremental as it builds on existing weakly-supervised approaches.
The paper tackled weakly-supervised semantic image segmentation by introducing a loss function based on seeding, expanding, and constraining principles, resulting in substantially better segmentations than previous state-of-the-art methods on the PASCAL VOC 2012 dataset.
We introduce a new loss function for the weakly-supervised training of semantic image segmentation models based on three guiding principles: to seed with weak localization cues, to expand objects based on the information about which classes can occur in an image, and to constrain the segmentations to coincide with object boundaries. We show experimentally that training a deep convolutional neural network using the proposed loss function leads to substantially better segmentations than previous state-of-the-art methods on the challenging PASCAL VOC 2012 dataset. We furthermore give insight into the working mechanism of our method by a detailed experimental study that illustrates how the segmentation quality is affected by each term of the proposed loss function as well as their combinations.