ICCV 2023 AV4D Workshop
MISAR: A Multimodal Instructional System with Augmented Reality
A multimodal instructional system with augmented reality that combines speech, vision and language for task guidance.
My part. Helped pivot the DARPA project toward vision-language and reasoning models via a video-to-text prototype (encoder-to-GPT-2).
A multimodal instructional system with augmented reality that combines speech, vision and language for task guidance.