Learning random-walk label propagation for weakly-supervised semantic segmentation
This addresses the problem of reducing annotation costs for semantic segmentation in computer vision, though it is incremental as it builds on existing label propagation and CNN methods.
The paper tackles the challenge of expensive training data for semantic segmentation by proposing a method that propagates sparse labels to generate dense labelings using random-walk hitting probabilities, which are learned jointly with a CNN segmentation network, resulting in improved performance over naive approaches.
Large-scale training for semantic segmentation is challenging due to the expense of obtaining training data for this task relative to other vision tasks. We propose a novel training approach to address this difficulty. Given cheaply-obtained sparse image labelings, we propagate the sparse labels to produce guessed dense labelings. A standard CNN-based segmentation network is trained to mimic these labelings. The label-propagation process is defined via random-walk hitting probabilities, which leads to a differentiable parameterization with uncertainty estimates that are incorporated into our loss. We show that by learning the label-propagator jointly with the segmentation predictor, we are able to effectively learn semantic edges given no direct edge supervision. Experiments also show that training a segmentation network in this way outperforms the naive approach.