Semantic Segmentation of Underwater Imagery: Dataset and Benchmark
This work addresses the problem of limited data for underwater robot vision, enabling improved scene understanding and autonomy, though it is incremental as it builds on existing segmentation methods.
The authors tackled the lack of large-scale datasets for underwater semantic segmentation by introducing SUIM, a dataset with over 1500 annotated images across eight categories, and benchmarked state-of-the-art methods, including SUIM-Net, which offers competitive performance with fast inference for robot applications.
In this paper, we present the first large-scale dataset for semantic Segmentation of Underwater IMagery (SUIM). It contains over 1500 images with pixel annotations for eight object categories: fish (vertebrates), reefs (invertebrates), aquatic plants, wrecks/ruins, human divers, robots, and sea-floor. The images have been rigorously collected during oceanic explorations and human-robot collaborative experiments, and annotated by human participants. We also present a benchmark evaluation of state-of-the-art semantic segmentation approaches based on standard performance metrics. In addition, we present SUIM-Net, a fully-convolutional encoder-decoder model that balances the trade-off between performance and computational efficiency. It offers competitive performance while ensuring fast end-to-end inference, which is essential for its use in the autonomy pipeline of visually-guided underwater robots. In particular, we demonstrate its usability benefits for visual servoing, saliency prediction, and detailed scene understanding. With a variety of use cases, the proposed model and benchmark dataset open up promising opportunities for future research in underwater robot vision.