search

llmpass / RSTT

Official pytorch implementation of paper "RSTT: Real-time Spatial Temporal Transformer for Space-Time Video Super-Resolution"
117 stars 20 forks source link

Train on myself datasets #4

Closed jackxx33 closed 2 years ago

jackxx33 commented 2 years ago

Hello,I am very interested in your work. I want to use this model to train my own dataset,is this possible?

llmpass commented 2 years ago

Definitely possible, just prepare the lmdb in the same way: 4 frame input 7 frame output.

jackxx33 commented 2 years ago

Does this model have to 4 frame input 7 frame output? Can I change some places to achieve 2x or 3x frame interpolation? And why the num_frames in 'RSTT-S.yml' is 7 instead of 4? Thanks!

llmpass commented 2 years ago

Yes. It can achieve 2x or 3x frame interpolation, but you have to change the code to feed the right number of output frames. Then, change the code starts at https://github.com/llmpass/RSTT/blob/9a27232d7066a545cd026d24b8c987e73e69c76f/models/RSTT.py#L145.