How to Train Monodepth2 on Custom Dataset made by image sequences
giuseppecialdella opened this issue · 0 comments
giuseppecialdella commented
Hello everyone,
i'm working with a MAV which navigates in an environment where it does collision avoidance and i want to implement Monodepth2.
I'm using Airsim and some custom code to collect image sequences, so i would like to use this module to extract DepthMaps but i never used Monodepth or any Depth Module/Algorithm before.
I've read the paper but i'm asking if there is a real tutorial to follow, starting only from images, not divided or modeled as a dataset.
Thanks