Trung-Hieu Le

National Alliance on Mental Illness, Hanoi University of Science and Technology

Papers

Total Citations

H-Index

About

Trung-Hieu Le is an emerging researcher in the field of human action recognition (HAR), with a focused interest in leveraging advanced deep learning architectures for multimodal sensor data. His work primarily addresses the challenge of accurately interpreting human movements from inertial sensors—such as those found in smartphones and wearables—for applications in healthcare monitoring, smart home systems, and human–robot interaction. Le’s major contributions include pioneering the use of Transformer models for HAR, as demonstrated in his 2022 paper on action recognition from inertial sensors, which explores how attention mechanisms can capture long-range dependencies in motion sequences. More recently, he introduced the Mamba-MHAR framework (2025), a novel multimodal approach that integrates Mamba state-space models with traditional deep learning to efficiently fuse data from multiple sensor modalities. Although his work is early-stage, with each paper accumulating 2 citations, the timeliness and innovation of his methods—particularly the application of state-space models to HAR—signal strong potential for future impact. Le’s research is notable for bridging cutting-edge sequence modeling techniques with practical, real-world sensing challenges, positioning him as a promising voice in the next wave of HAR advancements.