Anis Najar

Papers

Total Citations

H-Index

About

Anis Najar is a researcher specializing in human-robot interaction, interactive machine learning, and social robotics, with a focus on enabling robots to learn naturally and intuitively from human feedback. His most significant contribution lies in developing methods that allow robots to interpret unlabeled human instructions and evaluative signals without requiring rigid, predefined interaction protocols — a breakthrough that substantially improves the usability of interactive learning systems. Najar's most cited work, "Training a robot with evaluative feedback and unlabeled guidance signals" (2016, 25 citations), introduced a novel framework where feedback signals are dynamically mapped to reward values, enabling robots to simultaneously learn tasks and decode the meaning of human guidance. This line of research continued with "Interactively shaping robot behaviour with unlabeled human instructions" (2020, 15 citations), further refining flexible human-robot communication. His earlier work on Social-Task Learning (2015) demonstrated an innovative architecture using Learning Classifier Systems to concurrently model social interaction and accelerate task learning, reducing the burden on human trainers. Collectively accumulating nearly 50 citations, Najar's research meaningfully advances the accessibility and adaptability of robot training, making intelligent systems more responsive to natural, unconstrained human communication.