Real-time semantic scene completion via feature aggregation and conditioned prediction

Reported on 3 benchmarks across 3 tasks · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.