train with my own video

ClementPinard / SfmLearner-Pytorch

Pytorch version of SfmLearner from Tinghui Zhou et al.

MIT License

1.01k stars 224 forks source link

train with my own video #131

Closed ynjiun closed 3 years ago

ynjiun commented 3 years ago

Hi,

Great job!

Question: if I want to train the DispNetS only with my own video, do I need to have camera parameters for training? What would be the minimal requirements (for example, 3 sequence images) to train DispNetS?

Any suggestion on how to modify prepare_train_data.py to take the minimal input (presumably only the video needed) to generate the train dataset?

Thank you very much for your help.

ynjiun commented 3 years ago

I think you already answered my question in https://github.com/ClementPinard/SfmLearner-Pytorch/issues/108

issue closed!