Similarity-based representation learning combined with density-based aggregation could be a promising strategy for point cloud segmentation in highly granular particle detectors, according to a new study in PRX Intelligence: https://go.aps.org/4wlRI8C
