Yen-Yu Chang

Papers

Total Citations

H-Index

About

Yen-Yu Chang is a researcher at the forefront of multisensory object-centric learning and sim-to-real transfer in robotics and embodied AI. His work addresses a fundamental gap in the field: the lack of realistic, multisensory object representations for training intelligent agents. As the lead author of "ObjectFolder 2.0," Chang introduced a groundbreaking dataset that virtualizes objects with synchronized visual, acoustic, and tactile data, enabling robots to learn from rich, multimodal experiences in simulation before transferring to the real world. This work builds on the foundational ObjectFolder 1.0, which first demonstrated the potential of multisensory object modeling. Chang’s contributions are critical for advancing robotic manipulation, scene understanding, and interactive perception, with his research already garnering attention in top venues like CVPR. By creating more realistic digital twins, he is helping to bridge the reality gap that has long hindered sim-to-real transfer, making his work highly impactful for students and researchers exploring the intersection of computer vision, robotics, and multimodal learning.