Max Robotics — Global Service Robot Directory + US Certification

About

Nora Castner is a leading researcher at the intersection of computer vision, human-computer interaction, and assistive AI systems. Her work primarily focuses on leveraging gaze behavior to predict and prevent errors in AI-supported environments, with a strong emphasis on accessibility for individuals with motor impairments. Castner’s most cited paper, *“MAM: Transfer Learning for Fully Automatic Video Annotation and Specialized Detector Creation”* (2019, 49 citations), introduced a novel transfer learning framework that enables fully automatic video annotation and the creation of specialized object detectors—a foundational contribution to efficient computer vision pipelines. More recently, she has pioneered the use of eye-tracking to anticipate system errors in AI-assisted robotic arms, as demonstrated in her 2023 study on potential error prediction in simulated robot interactions and her 2024 VR-based work on communication breakdowns. These studies explore how gaze patterns can serve as early indicators of system failures, enhancing safety and trust in AI-driven mobility devices. Castner’s innovative integration of gaze analysis with virtual reality simulations represents a significant step toward more intuitive, error-resilient assistive technologies, making her a notable voice in accessible AI design.