no code implementations • CVPR 2023 • Camilo L. Fosco, SouYoung Jin, Emilie Josephs, Aude Oliva
We show that including information from the ETM during training improves action recognition and anticipation performance on various egocentric video datasets.
no code implementations • 1 Jun 2022 • Camilo Fosco, Emilie Josephs, Alex Andonian, Allen Lee, Xi Wang, Aude Oliva
Second, they allow us to generate novel "Deepfake Caricatures": transformations of the deepfake that exacerbate artifacts to improve human detection.