Pedro Victor Vieira de Paiva
Centro de Tecnologia da Informação Renato Archer, Universidade Estadual de Campinas (UNICAMP)
Papers
3
Total Citations
15
H-Index
3
About
Pedro Victor Vieira de Paiva is a robotics researcher dedicated to advancing human-robot interaction in unstructured, dynamic environments. His work centers on social perception, focusing on how robots can interpret human behavior using only lightweight, in-the-wild sensor data. In his highly cited 2020 paper, "Estimating Human Body Orientation using Skeletons and Extreme Gradient Boosting" (7 citations), he pioneered a method to infer body orientation from single-camera skeleton data—a critical cue for social path planning and approaching—without relying on expensive multi-sensory setups. That same year, he introduced ROSANA (5 citations), a mobile robot platform designed for social interaction, tackling the challenge of balancing high-level perception with limited onboard computational power. Most recently, in 2025, Paiva unveiled SkelETT—Skeleton-to-Emotion Transfer Transformer (3 citations), a novel architecture that recognizes human emotion purely from skeletal motion, bypassing traditional reliance on facial expressions or audio. By combining efficient machine learning with real-world deployability, Paiva’s contributions are shaping how socially interactive robots perceive and respond to people, making autonomous social navigation and affective computing more accessible and robust.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3SkelETT—Skeleton-to-Emotion Transfer Transformer3 citations · 2025