Carlos Fernandez‐Granda

Courant Institute of Mathematical Sciences

Papers

Total Citations

H-Index

About

Carlos Fernandez‑Granda is a leading researcher in machine learning for fine‑grained action recognition, with a particular focus on high‑temporal‑resolution analysis of human movement. His work bridges computer vision and healthcare, developing models that can identify subtle, sub‑second actions from video and kinematic data—a critical capability for applications in robotics, rehabilitation, and smart health. His most cited paper, “StrokeRehab: A Benchmark Dataset for Sub‑second Action Identification” (2022, 5 citations), introduced a novel benchmark that challenges conventional action‑recognition methods, which typically target coarse, longer‑duration activities. By providing a dataset of precise, rapid movements, Fernandez‑Granda enables the development of more sensitive and clinically relevant models. In his follow‑up work, “Sequence‑to‑Sequence Modeling for Action Identification at High Temporal Resolution” (2021, 2 citations), he advanced the state of the art by designing architectures capable of capturing the fine temporal structure of actions. Though his citation counts are modest, his contributions are foundational for emerging fields like stroke rehabilitation monitoring and assistive robotics. Fernandez‑Granda’s research is distinguished by its focus on temporal precision, pushing the boundaries of what machine learning can achieve in real‑time, high‑stakes environments.