search

facebookresearch / robust-dynrf

An algorithm for reconstructing the radiance field of a dynamic scene from a casually-captured video.
MIT License
219 stars 11 forks source link

poor results in DAVIS datasets #16

Open cv-lab-x opened 1 year ago

cv-lab-x commented 1 year ago

Hi, thanks for your great work, i have preprocess the davis datasets as you mentioned in readme, and use your config DAVIS.txt，but the results is poor, can't reproduce the demo you show in the readme? can you share your preprocessed DAVIS datasets?@alex04072000 Looking forward to your reply, thanks!

cuitao111 commented 1 year ago

depthvideo video I also get poor results in DAVIS datasets as shown in the gif above, and it takes a lot of time to train a short video clip, is there any way to speed up the training? @alex04072000

Pinocchio-Jia commented 10 months ago

Hi, thanks for your great work, i have preprocess the davis datasets as you mentioned in readme, and use your config DAVIS.txt，but the results is poor, can't reproduce the demo you show in the readme? can you share your preprocessed DAVIS datasets?@alex04072000 Looking forward to your reply, thanks!

When I try to train another scenario in the DAVIS dataset by modifying the datadir and expname in the DAVIS. txt file, an error message appears. Have you encountered it again? How to solve it？ 1 image