Graph-boosted convolutional neural networks for semantic segmentation

Abstract

This paper investigates the problem of weakly-supervised semantic segmentation, where image-level labels are used as weak supervision. Inspired by the successful use of Convolutional Neural Networks (CNNs) for fully-supervised semantic segmentation, we choose to directly train the CNNs over the oversegmented regions of images for weakly-supervised semantic segmentation. Although there are a few studies on CNNs-based weakly-supervised semantic segmentation, they have rarely considered the noise issue, i.e., the initial weak labels (e.g., social tags) may be noisy. To cope with this issue, we thus propose graph-boosted CNNs (GB-CNNs) for weakly-supervised semantic segmentation. In our GB-CNNs, the graph-based model provides the initial supervision for training the CNNs, and then the outcomes of the CNNs are used to retrain the graph-based model. This training procedure is iteratively implemented to boost the results of semantic segmentation. Experimental results demonstrate that the proposed model outperforms the state-of-the-art weakly-supervised methods. More notably, the proposed model is shown to be more robust in the noisy setting for weakly-supervised semantic segmentation.