EPFL-VILAB / omnidata

A Scalable Pipeline for Making Steerable Multi-Task Mid-Level Vision Datasets from 3D Scans [ICCV 2021]
Other
395 stars 49 forks source link

Train own dataset #62

Open rick-dn opened 1 month ago

rick-dn commented 1 month ago

My dataset has got rgb images and their corresponding depth images. Can I train my dataset on this model?

YacineDeghaies commented 3 weeks ago

@rick-dn were you able to train on your own dataset ?