You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have built the extensions and then attempt to run a demo of sfm_bench provided at the readme. But after several tries, I always get an error with the following message:
ii, jj = torch.as_tensor(es, device=self.device).unbind(dim=-1)
ValueError: not enough values to unpack (expected 2, got 0)
An interesting observation is the images of the sfm_bench demo could be shown usually. While, when entering into the driod.terminate() function, the program crashes with the above error. I find that the value of delta.norm(dim=-1).mean().item() (in motion_filter.py) is strange ( ~0.0027 ). As the filter_threshold is set at about 2.4, I guess the frame data can not be added into self.video
My question is does it only happen on windows? I will give try on ubuntu to take a further test. Thx for reply.
The text was updated successfully, but these errors were encountered:
I have built the extensions and then attempt to run a demo of
sfm_bench
provided at the readme. But after several tries, I always get an error with the following message:An interesting observation is the images of the
sfm_bench
demo could be shown usually. While, when entering into thedriod.terminate()
function, the program crashes with the above error. I find that the value ofdelta.norm(dim=-1).mean().item()
(inmotion_filter.py
) is strange ( ~0.0027 ). As the filter_threshold is set at about 2.4, I guess the frame data can not be added intoself.video
My question is does it only happen on windows? I will give try on ubuntu to take a further test. Thx for reply.
The text was updated successfully, but these errors were encountered: