search

uzh-rpg / rpg_e2depth

Code for Learning Monocular Dense Depth from Events paper (3DV20)
GNU General Public License v3.0
113 stars 31 forks source link

About MVSEC sequence #1

Closed wl082013 closed 3 years ago

wl082013 commented 3 years ago

Hi, nice work!

Could you also share the sequences of the MVSEC dataset that you used to train and test the network? I mean the samples you mentioned in the paper, where the training split was about 8523 samples and the test split was about 1826 samples. It would be a great help for real-world data test using the sequence you used.

Another question is that you use left or right camera in the MVSEC dataset?

Thanks a lot.

jhidalgocarrio commented 3 years ago

I just added the link to the MVSEC dataset we used. We used DAVIS left camera. Please have a look at the website: rpg.ifi.uzh.ch/e2depth