LMC-Memory
LMC-Memory copied to clipboard
Official PyTorch implementation of "Video Prediction Recalling Long-term Motion Context via Memory Alignment Learning" (CVPR 2021 Oral)
Question: I typed 10 frames to predict 20 frames, so short_len was set to 10, long_len to 20, and out_len to 20, but I ran into problems and couldn't train
I try my own dataset but find out that the outputs are 1-channel pictures, but your paper use Human 3.6M dataset and can work and print RGB pictures, I'm confirmed...
How to convert the downloaded datasets into frames and store it according to your requirements? Or, can you provide the download link of the datasets you have processed?
I try to run the model using movingmnist but failed, it shows "RuntimeError: The size of tensor a (7) must match the size of tensor b (4) at non-singleton dimension...
Hi ! I just want to point something kind of special, so it seems that your code even if I specified shuffle=True, don't make the valid dataloader shuffle. So the...