Ego-4D
Emerging4papers using it
2025first seen
The 'Ego-4D' dataset is a benchmark that contains egocentric video data used to evaluate models for predicting human actions and their visual outcomes in the context of human-object interactions.
Papers using Ego-4D (4)
- V-JEPA 2.1: Unlocking Dense Features in Video Self-Supervised LearningEgo-centric Predictive Model Conditioned on Hand TrajectoriesRoboPearls: Editable Video Simulation for Robot ManipulationTASTE-Rob: Advancing Video Generation of Task-Oriented Hand-Object Interaction for Generalizable Robotic Manipulation