Abstract
In situated dialogue, listeners resolve referring expressions incrementally and their gaze often attends to the described objects in the context. We have looked at how listener gaze compares to a statistical reference resolution model that works incrementally. We found that listeners gaze at referred objects even before a referring expression begins, suggesting that salience and prior information is important in reference resolution models.
Original language | American English |
---|---|
State | Published - Aug 2014 |
Externally published | Yes |
Event | RefNet Workshop on Computational and Psychological Models of Reference Comprehension and Production - Edinburgh, Scotland Duration: 1 Aug 2014 → … |
Conference
Conference | RefNet Workshop on Computational and Psychological Models of Reference Comprehension and Production |
---|---|
Period | 1/08/14 → … |
EGS Disciplines
- Computer Sciences