codeslake / realtime_vdblr Goto Github PK

Was anyone able to run the code on your own frames? Having a Colab Notebook for that would be so much useful. Why are code based on papers are so poorly documented or either impossibly hard to follow. I am not a PhD student and don't need to replicate your results, just give me please instructions on how can I run it on my own frames or video.

Rank[-1]: _IncompatibleKeys(missing_keys=['module.Network.deblur_layer.0.weight', 'module.Network.deblur_layer.0.bias'], unexpected_keys=[])

I trained a model.but it is diffrent from your source model.Can you help me?

Could the codebase here be ran on CPU instead? If so, what are the modifications that need to be done?

OSError: CUDA_HOME environment variable is not set. Please set it to your CUDA install root.

I cannot find the inference code for video. Could you provide the code?

This liteFlowNet.pytorch pre-trained model is not found in the provided Google drive link

May I ask if the indicators of PSNR in the paper are for the training set or the validation set?