stonemo / eva-gcn Goto Github PK

We tried reproducing your code for a head pose estimation study. But we cannot get the code running since it does not have the files which has been used to create the datasets or the files does not contain the datasets which has been used to train the model. Can we plz get the dataset creation files or the datasets which has been used to train the model. It would be a great help to get your code running and compare your work against ours. Thank you very much in advance.!!

Thanks for your great work and for sharing the codes! May I ask you why you use ST-GCN? Your case seems to be an S-GCN. Where did you use the temporal dimension?

I'm trying EVA-GCN in colabs. Getting following error. Any idea to solve this?

Traceback (most recent call last):
File "main.py", line 29, in
p = Processor(sys.argv[2:])
File "/content/EVA-GCN/processor/processor.py", line 30, in init
self.init_environment()
File "/content/EVA-GCN/processor/processor.py", line 39, in init_environment
super().init_environment()
File "/content/EVA-GCN/processor/io.py", line 63, in init_environment
torchlight.occupy_gpu(gpus)
File "/content/EVA-GCN/torchlight/gpu.py", line 35, in occupy_gpu
torch.zeros(1).cuda(g)
RuntimeError: CUDA error: invalid device ordinal