Benjamin Stoler
Papers
1
Total Citations
2
H-Index
1
About
Benjamin Stoler is a rising researcher at the intersection of computer vision, robotics, and human-robot interaction, with a focused interest in pedestrian trajectory prediction from egocentric perspectives. His most notable contribution, the paper "T2FPV: Dataset and Method for Correcting First-Person View Errors in Pedestrian Trajectory Prediction" (2023), addresses a critical gap in socially-aware robotics: most trajectory prediction models rely on third-person, bird’s-eye views, but real-world robots operate from a first-person, egocentric viewpoint. Stoler introduced both a novel dataset and a correction method specifically designed to handle the unique errors—such as perspective distortion and occlusion—that arise in first-person view (FPV) settings. This work is foundational for developing robots that can safely and naturally navigate crowded spaces alongside humans. Though early in his career, with 2 citations to date, his research is gaining traction as the field increasingly prioritizes embodied AI and real-world deployment. Stoler’s contributions are particularly relevant for students and researchers working on autonomous navigation, social robotics, and vision-based prediction, offering a practical bridge between theoretical models and the messy, dynamic realities of human environments.
Research Focus
Key Achievements
Top Papers
- 1