Cecilia Curreli to Present EquiFusion at ECCV
She describes equivariant latent diffusion models unlocking multi-dataset training for human motion prediction.
Cecilia Curreli stated on X that she will present EquiFusion at ECCV next week. The research covers equivariant latent diffusion models and their use in training and inference on multiple datasets for human motion prediction. She offered to discuss the topic with interested attendees at the conference. The post included an animated diagram showing past inputs leading to multiple future outputs labeled EquiFusion.
Combined views
893
2 posts, first seen 1d ago