RDPN

Pretrained checkpoints and metadata for RDPN. Training data archives are in NTU-AI2Lab/RDPN-data.

Contents

• checkpoints/lmso/ — per-object checkpoints for LM (13 objects)
• checkpoints/lmoso/ — per-object checkpoints for LM-O (8 objects)
• checkpoints/mp6dso/ — per-object checkpoints for MP6D (obj_01–obj_20)
• metadata/ — models, image sets, bounding boxes and test targets for lm, lmo, MP6D and YCBV

Download

hf download NTU-AI2Lab/RDPN --local-dir RDPN

or a single checkpoint:

from huggingface_hub import hf_hub_download
path = hf_hub_download("NTU-AI2Lab/RDPN", "checkpoints/lmso/ape.pth")