Teaching Language to Deaf Infants with a Robot and a Virtual Human
summary
An integrated system combines a 3D-printed Maki robot that cues gaze and attention with a virtual human from photogrammetry and motion capture that produces real signer gestures, linked to a perception stack using a commercial eye tracker and thermal camera. Over three years 68 infants test iterative versions; a 13‑month deaf infant engages up to a minute, an 8‑month imitates a robot gesture, an 11‑month warms to the system when parent joins. Results show reliable engagement but no demonstrated learning or generalization from the short (~3 minute) sessions.
AI-written summary · about
chapters (13)
- 0:00Introduction to project
- 1:56Why robots and virtual humans
- 4:47Team and project goals
- 6:34System design overview
- 7:17Robot design modifications
- 9:06Virtual human creation
- 10:31Iterative testing process
- 11:14First case study Albert
- 12:17Second case study Bella
- 13:03Third case study Celia
- 15:55Key findings summary
- 17:19Limitations and future work
- 18:04Q&A session
related talks
16mSomeone to Read with: Design of and Experiences with an In-Home Learning Companion Robot for Reading
28mPreference Learning from Minimal Human Feedback for Interactive
08mA Simple Nod of the Head: The Effect of Minimal Robot Movements on Children's Perception of a ...
31mArtificial Intelligence for Artificial Humanoids
10mA Novel Architectural Method for Producing Dynamic Gaze Behavior in Human-Robot Interactions
10mInclusive’R’Stories: An Inclusive Storytelling Activity with an Emotional Robot
06mInterview with Jong Lee
10mIndividual Differences of Children with Autism in Robot-Assisted Autism Therapy