Sepehr Valipour

University of Tehran, University of Alberta

Papers

Total Citations

H-Index

About

Sepehr Valipour is a robotics and artificial intelligence researcher whose work sits at the intersection of deep learning, computer vision, and human-robot interaction. His most significant contributions focus on bridging the gap between the impressive performance of Convolutional Neural Networks on benchmark datasets and their practical deployment in real-world robotic systems. Most notably, Valipour has pioneered incremental learning frameworks for robot perception, enabling robots to continuously acquire new visual knowledge through human-robot interaction rather than relying solely on fixed, pre-trained models — a breakthrough that has earned his 2017 work 34 citations and established him as a meaningful voice in adaptive robot learning. Beyond perception and learning, Valipour has demonstrated breadth in his research through applied UAV systems engineering, contributing a vision-based power line tracking system for autonomous aerial inspection vehicles — a technically demanding problem combining image processing and real-time control. His graduate thesis, *Deep Learning in Robotics*, further reflects his commitment to diagnosing and overcoming the systemic barriers that prevent cutting-edge machine learning from translating into reliable robotic applications. Across his body of work, Valipour consistently pursues the practical realization of intelligent, adaptable robotic systems capable of operating meaningfully in dynamic human environments.