Nikola Raicevic

University of Rochester, University of California San Diego

Papers

Total Citations

H-Index

About

Nikola Raicevic is pioneering the frontier of human-robot interaction by teaching machines to understand the nuanced, time-sensitive language we use every day. His research centers on natural language grounding in dynamic, cluttered environments—a critical challenge for robots operating in homes and offices alongside humans. Raicevic’s major contributions include developing a language-guided, temporally adaptive perception system that allows robots to efficiently interpret instructions by focusing only on relevant visual information, dramatically improving performance in real-world settings. His 2023 paper on this approach has already garnered 4 citations, signaling its foundational impact. In his most recent 2024 work, Raicevic tackles an even more complex problem: enabling robots to ground language that refers to the past, present, or future state of objects. This breakthrough gives robots a temporal understanding of how spatial relationships evolve, allowing them to follow instructions like “the cup that was on the table” or “the box you will move later.” By bridging the gap between human temporal reasoning and robotic perception, Raicevic is laying the groundwork for truly collaborative, intuitive human-robot partnerships.