In order to evaluate and rank the performances of the participant methods, we use 3D topology-based segmentation metrics. Each metric has an updated leader-board.

The metrics are:

  • Adapted Rand error: defined as 1 - the maximal F-score of the¬†Rand index (excluding the zero component of the original labels).
  • Other metrics to be announced.


You can calculate the metric values on the training data yourself using any of the following scripts: