Learned Watershed: End-to-End Learning of Seeded Segmentation
This work improves segmentation accuracy for biomedical imaging, but it is incremental as it builds on existing watershed methods.
The paper tackles the problem of seeded segmentation by training watershed computation jointly with boundary map prediction, achieving state-of-the-art results on the CREMI segmentation challenge.
Learned boundary maps are known to outperform hand- crafted ones as a basis for the watershed algorithm. We show, for the first time, how to train watershed computation jointly with boundary map prediction. The estimator for the merging priorities is cast as a neural network that is con- volutional (over space) and recurrent (over iterations). The latter allows learning of complex shape priors. The method gives the best known seeded segmentation results on the CREMI segmentation challenge.