Scene Change Detection on Semantic Scene Understanding Challenge (passive actuation & ground-truth localisation)

Metric: avg_label (higher is better)