OSMO: Open-vocabulary Self-eMOtion Tracking
OSMO introduces egocentric self-emotion tracking with a large-scale dataset, a multi-task benchmark, and OSIRIS, a multimodal model for coherent emotion timelines.
mohamed-abdelfattah
•
•
1 min read
OSKAR: Omnimodal Self-supervised Knowledge Abstraction and Representation
OSKAR is a self-supervised multimodal foundation model that learns in the latent space by predicting masked multimodal features.
mohamed-abdelfattah