Data Science Talk Series 

Title: Data Quality-aware Graph Machine Learning

Abstract:

Abstract: In this talk, I will introduce a new type of human-computer interface - Mobile Acoustic Field (MAF). MAF explores a variety of onboard sensors such as speaker transducers, IMU, and feedforward&feedback microphones on earphones/earbuds to capture a wide range of acoustic activities both inside-out and outside-in. I anticipate that the exponential advancement of generative AI, supported by the sheer size of cloud computing resources, will enable personal agents to sense, understand, and further interact with this mobile acoustic field, offering many exciting mobile services to users. My research group has been working on this research field for over three years. In this talk, I will summarize our preliminary results, discuss the downstream applications, technical challenges, and our vision on this exciting research field.