Closed mengW6 closed 2 months ago
Hi, the training and inference processes are mainly in mld/models/modeltype/mld.py. The denoiser network is in mld/models/architectures/mld_denoiser.py.
More specifically, code related to style extraction is here, where we use the pretrained MotionCLIP to encode style motion. Code related to trajectory feature extraction is here, where we use a transformer-based network to encode trajectory.
If you have any other questions, feel free to ask. :)
Thank you for your reply :)
hello, thank you for your great works! May I know which file contains the operations related to style extraction and trajectory feature extraction?