Narrowing the Coordinate-Frame Gap in Behavior Prediction Models
Abstract
We study scene-centric motion forecasting and distillation techniques to improve both accuracy and efficiency in behavior prediction models.
Type
Publication
Proceedings of the IEEE International Conference on Robotics and Automation
Rami Al-Rfou
Member of Technical Staff - TLM
My research interests include language modeling, embodied AI, motion forecasting, and multilingual modeling.