Yuuna Hoshi

Georgia Institute of Technology

Papers

4

Total Citations

102

H-Index

3

About

Yuuna Hoshi is a leading researcher in assistive robotics, with a primary focus on enabling people with disabilities to perform activities of daily living (ADLs) through intelligent robotic systems. Her major contributions center on developing robust, multimodal anomaly detection and execution monitoring frameworks for robot-assisted feeding—a critical ADL for independent living. Hoshi pioneered the use of LSTM-based variational autoencoders to fuse high-dimensional, heterogeneous sensory data, allowing robots to detect and classify a wide range of anomalies during real-world assistive tasks. Her most cited work, "A multimodal execution monitor with anomaly classification for robot-assisted feeding" (60 citations), established foundational methods for safe, semi-autonomous manipulation. She further advanced the field by designing and evaluating general-purpose mobile manipulators for active feeding, demonstrating practical deployment in real-world settings. Hoshi also contributed to robot vision through multitask learning for object instance recognition and 3D pose estimation, with adaptive loss balancing to improve accuracy. Her work bridges the gap between complex perception, anomaly detection, and safe physical assistance, directly impacting the quality of life for individuals with disabilities. With over 100 total citations, Hoshi’s research continues to shape the future of assistive manipulation and human-robot interaction.

Research Focus

Key Achievements

3
H-Index
4
Papers
102
Total Citations
26
Avg Citations/Paper
🏆 Most Cited Paper
A multimodal execution monitor with anomaly classification for robot-assisted feeding
60 citations · 2017
📈 Most Prolific Year: 2019 (2 Papers)
🤝 Key Collaborators: 11
🏛 Institutions: Georgia Institute of Technology

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago