Max Robotics — Global Service Robot Directory + US Certification

About

Nicholas Ruozzi is a leading researcher in robot perception and computer vision, with a primary focus on enabling robots to perceive and interact with objects they have never seen before. His work centers on the critical challenge of **unseen object instance segmentation**—the ability to identify and separate novel objects in cluttered environments without prior training. Ruozzi’s major contributions include pioneering **self-supervised learning frameworks** that allow robots to improve their segmentation skills through long-term physical interaction, such as pushing and grasping objects. He also developed the **Mean Shift Mask Transformer**, a novel architecture that integrates classical mean shift clustering with modern transformer models for robust, real-time segmentation. With over 29 citations across his most-cited works, his research has immediate applications in robotic manipulation and automation. Notably, his 2023 work on long-term robot interaction demonstrates a practical pathway for continuous learning in the wild, while his 2025 NIDS-Net framework extends his expertise to few-shot instance detection and segmentation. Ruozzi’s work is essential reading for anyone interested in bridging the gap between computer vision and autonomous robotic systems.