BROWSE
Detail
Conference Paper
가이드 독을 위한 사용자 인터랙션
Cited - time in 
Share 
Share
- Issue Date
- 2024-06
- Citation
- 대한전자공학회 학술 대회 (하계) 2024, pp.3317-3320
- Publisher
- 대한전자공학회
- Language
- Korean
- Type
- Conference Paper
- Abstract
- This paper explains the configuration and implementation of the user interaction module in a guidance robot system designed for visually impaired individuals. The module comprises an internal and external situation recognition part, a situation explanation part, and a user voice recognition part. Currently, the implementation has been completed up to the static object detection function in the external situation recognition part with the YOLOv8 algorithm. The dataset of 7,493 images across 47 routes was collected and annotated, which was then divided into training and evaluation sets. Experimental results indicate that static object detection is rapid and highly accurate, validating its application in an external situation recognition part of the interaction module.
- KSP Keywords
- Guidance robot, Highly accurate, Internal and external, Situation recognition, Training and evaluation, User interaction, Visually impaired individuals, Voice Recognition, object detection, robot system