Papers
2
Total Citations
13
H-Index
2
About
Eri Onami is a researcher advancing the frontier of embodied AI and human-robot interaction, with a primary focus on first-person perception and language grounding. Her key research area centers on developing agents that can understand and act upon natural language instructions from an egocentric viewpoint—a critical capability for smart glasses and autonomous robots. Onami’s major contribution is the creation of **RefEgo**, the first large-scale dataset for referring expression comprehension (REC) sourced from the Ego4D corpus. This work addresses the challenging task of grounding textual expressions to objects in dynamic, first-person scenes, moving beyond traditional third-person benchmarks. The RefEgo dataset, published in 2023, has already garnered 11 citations, signaling its immediate impact on the computer vision and robotics communities. By enabling agents to localize objects based on intuitive text commands from a wearer’s perspective, Onami’s research directly supports the development of context-aware, assistive technologies. Her work stands as a foundational step toward more natural human-machine collaboration in real-world environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2