Learning to Detect Semantic Boundaries with Image-level Class Labels
This work addresses the problem of reducing annotation costs for semantic boundary detection in computer vision, though it is incremental as it builds on existing classification networks and MIL strategies.
The paper tackles semantic boundary detection using only image-level class labels for supervision, formulating it as a multiple instance learning problem and training a model that generates pseudo labels, achieving competitive performance on the SBD dataset compared to methods with stronger supervision.
This paper presents the first attempt to learn semantic boundary detection using image-level class labels as supervision. Our method starts by estimating coarse areas of object classes through attentions drawn by an image classification network. Since boundaries will locate somewhere between such areas of different classes, our task is formulated as a multiple instance learning (MIL) problem, where pixels on a line segment connecting areas of two different classes are regarded as a bag of boundary candidates. Moreover, we design a new neural network architecture that can learn to estimate semantic boundaries reliably even with uncertain supervision given by the MIL strategy. Our network is used to generate pseudo semantic boundary labels of training images, which are in turn used to train fully supervised models. The final model trained with our pseudo labels achieves an outstanding performance on the SBD dataset, where it is as competitive as some of previous arts trained with stronger supervision.