Radar signals have the potential to reveal information about the geometry and material composition of real world objects that usual image/video signals may lack, especially with the recent development of high-frequency radar chips (60GHz with 7GHz bandwidth). Specific objects or environment layouts have different radar signatures. The research will build a database of radar signatures from known objects. With the primitives from this database, we will attempt to predict characteristics of new objects based on their measured radar signatures. If appropriate, we may introduce joint video-radar signal processing. The research hopes to explore the feasibility of using machine learning, signal processing, and radar signatures from known environments and objects to classify new ones.

I am a senior studying Computer Science. Although I have experience relevant to my project, this research is unlike anything I’ve done before. I hope to use my knowledge of artificial intelligence and signal processing to understand how these subjects are leveraged in a rigorous research setting. I’m very excited that the project has the potential to discover some novel applications of recently developed radar technology.