Yue Yang

I am currently a first-year Computer Science PhD student at The University of North Carolina at Chapel Hill (UNC-CH) , under the co-supervision of Prof. Daniel J. Szafir and Prof. Gedas Bertasius. Prior to embarking on this academic journey, I pursued my Master’s degree in Computer Science at the Georgia Institute of Technology (GaTech). During my time at GaTech, I worked closely with Prof. Matthew Gombolay. I was fortunate to collaborate with Prof. Greg Turk as well. It’s a true privilege to have worked with these outstanding academics.

My research interests lie in Robotics, Machine Learning (ML), and their application in Human-Robot Interaction (HRI). My desired research directions include but not limited to:

• Human-multi-robot interaction. Exploring human-multi-robot interaction frameworks that can promote human working efficiency.
• Human-involved data enhancement. 1). Leveraging multi-modalities (e.g., video contents, wearable sensors, natural language, etc.) to enable the robot to learn a human cognitive model from several sources instead of only observation-action tuples. 2). Further developing a learning framework that enables the robot to perform a task optimally given cross-domain demonstrations (i.e., demonstrations generated by an expert agent with different morphology or embodiment).
• Safe robot learning and deployment. 1). Investigating methods for learning human’s cognitive model of safety or risk via active inquiries so that human users can define these concepts on their own. 2). Examining the use of pre-learned generative models for human motion prediction.
• Learning from demonstrations. Investigating any design and implementation of novel LfD algorithms.
• Generative model applications. Utilizing generative models (e.g., VAE, GANs, Diffusion models, etc.) in a variety of interesting scenarios, such as image generation, video generation, motion prediction, etc.

I believe in Slow Science.