Mutual gaze detection and estimation: Towards human-robot interaction

Abstract:

Mutual gaze is an important part of social interaction and the perception of others emotional states and essential for establishing joint attention. It has been described as “the most powerful mode of developing a communicative link between humans”. Although gaze estimation and prediction using machine learning and computer vision is extensively studied in images and videos for automation, there is a lack of research in dialogue and interaction. In this study, we estimate gaze using a PyTorch-based model with the goal of implementing human-like mutual gaze in a robot.

Research areas:

• Dialogue
• Natural Language Processing

Year:

2022

Type of Publication:

In Proceedings

Book title:

Proceedings of the 26th Workshop On the Semantics and Pragmatics of Dialogue (DubDial) - Poster Abstracts

Pages:

272-274

ISSN:

2308-2275