Miriam Bilac

Papers

Total Citations

104

H-Index

About

Miriam Bilac is a leading researcher in human-robot interaction (HRI), with a primary focus on making conversations between people and social robots more natural and intuitive. Her key research areas include user engagement modeling, non-verbal communication cues, and dialogue management for autonomous systems. Bilac’s most impactful contribution is the creation of the UE-HRI dataset (2017, 95 citations), a unique resource capturing spontaneous, unscripted interactions between humans and the robot Pepper. This dataset, featuring 54 freely available interactions, has become a foundational tool for the HRI community, enabling researchers to study authentic engagement patterns rather than staged behaviors. In related work on gaze and filled pause detection (2017, 9 citations), she addressed a critical bottleneck in voice interfaces: the need for robots to recognize human conversational signals like hesitations and eye contact to avoid interrupting or dominating the dialogue. By developing algorithms that allow robots to detect when a person is thinking or about to speak, Bilac has advanced the goal of smooth, turn-taking conversations—moving beyond the rigid, “robotic” flows common in commercial assistants. Her work directly supports the design of more empathetic and socially aware robots, with lasting implications for assistive technology and collaborative AI.