NVlabs / FoundationPose

[CVPR 2024 Highlight] FoundationPose: Unified 6D Pose Estimation and Tracking of Novel Objects
https://nvlabs.github.io/FoundationPose/
Other
1.01k stars 112 forks source link

ycbv dataset confusion #111

Open cheny110 opened 3 weeks ago

cheny110 commented 3 weeks ago

When I run model free few-shot version on ycbv dataset, I get confused on the dataset pathes passed to the param. What's the different between YCB-Video and YCB_Video?

wenbowen123 commented 3 weeks ago

where are you referring to?

cheny110 commented 3 weeks ago

In Official README file, Run on public datasets (LINEMOD, YCB-Video) section, the given example command for model-free running is:

python bundlesdf/run_nerf.py --ref_view_dir /mnt/9a72c439-d0a7-45e8-8d20-d7a235d02763/DATASET/YCB_Video/bowen_addon/ref_views_16 --dataset ycbv

python run_ycb_video.py --ycbv_dir /mnt/9a72c439-d0a7-45e8-8d20-d7a235d02763/DATASET/YCB-Video --use_reconstructed_mesh 0

So what's the difference between DATASET/YCB_Video/ and DATASET/YCB-Video