Pasquale Coscia
University of Padua, University of Campania "Luigi Vanvitelli", Civita
Papers
6
Total Citations
170
H-Index
6
About
Pasquale Coscia is a researcher specializing in human motion forecasting, trajectory prediction, and intelligent systems for autonomous vehicles and social robotics. His work sits at the intersection of deep learning, computer vision, and human behavior understanding, with a particular focus on enabling machines to anticipate and respond to human movement in complex, real-world environments. Among his most influential contributions is his 2022 paper on Goal-driven Self-Attentive Recurrent Networks for trajectory prediction, which has garnered 61 citations and advances multi-modal forecasting by incorporating destination awareness and attention mechanisms. His earlier work on long-term path prediction using circular distributions (2017, 45 citations) demonstrated an innovative probabilistic approach to urban navigation challenges. He further extended multi-modal trajectory modeling through AC-VRNN (2021, 36 citations), an attentive conditional variational recurrent architecture designed for crowded scenario forecasting. Coscia has also contributed to pedestrian intent prediction and knowledge distillation for action anticipation, broadening his impact across autonomous driving and human-robot interaction. With over 170 cumulative citations, his research consistently addresses safety-critical applications, making his work highly relevant to students and practitioners developing next-generation intelligent transportation and robotic systems.
Research Focus
Key Achievements
Top Papers
- 1Goal-driven Self-Attentive Recurrent Networks for Trajectory Prediction61 citations · 2022
- 2Long-term path prediction in urban scenarios using circular distributions45 citations · 2017
- 3AC-VRNN: Attentive Conditional-VRNN for multi-future trajectory prediction36 citations · 2021
- 4Early Pedestrian Intent Prediction via Features Estimation11 citations · 2022
- 5Knowledge Distillation for Action Anticipation via Label Smoothing10 citations · 2021
- 6Social and Scene-Aware Trajectory Prediction in Crowded Spaces7 citations · 2019