Jyun-Ting Song

I am a second-year master's student at the Robotics Institute, Carnegie Mellon University, advised by Prof. Kris Kitani. Previously, I earned my B.S. in Electrical Engineering from National Taiwan Normal University, where I was advised by Prof. Jacky Baltes.

My research interest lies in the intersection of Computer Vision, Machine Learning and Robotics. Currently, I am working on human pose estimation and human-object contact detection. Previously, I have worked on humanoid control in both simulation and real-world robotics.

I love humans, and I am interested in understanding them. I hope my reseach can contribute to this direction.

Email  /  CV  /  Scholar  /  GitHub  /  Linkedin

profile photo

Publications

Harmony4D: A Video Dataset for In-The-Wild Close Human Interactions
Rawal Khirodkar*, Jyun-Ting Song*, Jinkun Cao, Zhengyi Luo, Kris Kitani
Neural Information Processing Systems (NeurIPS), 2024

A large-scale multihuman dataset captured in in-the-wild environments, featuring diverse dynamic activities

paper | abstract | project | dataset | bibtex |

Reinforcement Learning and Action Space Shaping for a Humanoid Agent in a Highly Dynamic Environment
Jyun-Ting Song, Guilherme Christmann, Jaesik Jeong, Jacky Baltes
Springer's Studies in Computational Intelligence, 2023

RL algorithm structure based on Proximal Policy Optimization (PPO) to train a humanoid agent to play a balance board in Isaac Gym

paper | abstract | project | bibtex

The CORSMAL benchmark for the prediction of the properties of containers
Alessio Xompero, et al.
IEEE Access, 2022

Estimate mass, type, and fill level of containers using multimodal dataset (visual, audio)

paper | abstract | project | bibtex |


adapted from Jon Barron's awesome webpage