Unsupervised Image Semantic Segmentation through Superpixels and Graph Neural Networks
This addresses the problem of image segmentation without labeled data for applications in scenarios with scarce data availability, representing an incremental advancement by integrating existing techniques in a new way.
The paper tackles unsupervised image semantic segmentation by proposing a novel approach combining Mutual Information Maximization, Neural Superpixel Segmentation, and Graph Neural Networks in an end-to-end manner, achieving improved accuracy over state-of-the-art methods on four popular datasets.
Unsupervised image segmentation is an important task in many real-world scenarios where labelled data is of scarce availability. In this paper we propose a novel approach that harnesses recent advances in unsupervised learning using a combination of Mutual Information Maximization (MIM), Neural Superpixel Segmentation and Graph Neural Networks (GNNs) in an end-to-end manner, an approach that has not been explored yet. We take advantage of the compact representation of superpixels and combine it with GNNs in order to learn strong and semantically meaningful representations of images. Specifically, we show that our GNN based approach allows to model interactions between distant pixels in the image and serves as a strong prior to existing CNNs for an improved accuracy. Our experiments reveal both the qualitative and quantitative advantages of our approach compared to current state-of-the-art methods over four popular datasets.