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
A question to the inference times, can your transformer be used for real-time inference frame by frame prediction, e.g. with a rtx 3090?
What about the training times, how many GPU hours are necessary?
Thanks in advance.
The text was updated successfully, but these errors were encountered:
Hi, the inference time depends on the resolution of the input mesh. You may find a detailed analysis in Figure 5 (page 6) of our paper. The training times also depends on the datasets. In all the datasets we tested, a 120hrs training is enough.
Thanks for this impressive results and work!!!
A question to the inference times, can your transformer be used for real-time inference frame by frame prediction, e.g. with a rtx 3090?
What about the training times, how many GPU hours are necessary?
Thanks in advance.
The text was updated successfully, but these errors were encountered: