MetricAnything: Scaling Metric Depth Pretraining with Noisy Heterogeneous Sources
Open paper
Benchmark
Model
Rank
Results
monocular-depth-estimation-on-ibims-1-metric
MetricAnything Student-PointMap
#5
AbsRel: 0.113
robot-manipulation-on-libero
MetricAnything (VLA depth distillation)
#128
Success Rate: 87.7
Rank counts only results with a code link.