Max Robotics — Global Service Robot Directory + US Certification

About

Laslo Dinges is a leading researcher at the intersection of computer vision, affective computing, and human-robot interaction (HRI). His work focuses on enabling robots to perceive, interpret, and respond to human emotional and behavioral cues, making interactions more intuitive and natural. Dinges’s most significant contribution is **L2CS-Net**, a pioneering deep learning architecture for fine-grained gaze estimation in unconstrained environments. With 105 citations, this work has become a key reference for appearance-based gaze tracking, critical for applications in virtual reality and collaborative robotics. Beyond gaze, Dinges has advanced **multimodal emotion recognition**, developing models that simultaneously predict discrete emotion categories and continuous valence/arousal dimensions, validated across challenging datasets like AffectNet and AFEW. He has also pioneered **sentiment-based engagement strategies** for robots, using real-time facial expression analysis to adapt robotic behavior and reduce user discomfort in industrial HRI scenarios. His research on semantic-aware environment perception further enables mobile robots to understand their surroundings contextually, facilitating more complex, human-centered cooperation. Through his integrated approach—combining fine-grained visual analysis with affective state prediction—Dinges is shaping the future of socially intelligent robots capable of safe, empathetic, and effective collaboration with humans.