-
Model trained with setting: 1 house 4 targets:
-
Fine-tune the model trained above with Policy Gradient.
-
Model trained with setting: 4 houses 4 targets; and tested in 1 house 4 targets.
-
Train a Mapper with VPN with MatterPort3D dataset.
Model trained with setting: 1 house 4 targets:
Fine-tune the model trained above with Policy Gradient.
Model trained with setting: 4 houses 4 targets; and tested in 1 house 4 targets.
Train a Mapper with VPN with MatterPort3D dataset.