Related Experiment Video
Updated: Sep 18, 2025
06:37
Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
Published on: December 15, 2023
4.1K
Toward a Recognition System for Mexican Sign Language: Arm Movement Detection
Gabriela Hilario-Acuapan1, Keny Ordaz-Hernández1, Mario Castelán1
1Robotics and Advanced Manufacturing Department, Centre for Research and Advanced Studies (CINVESTAV), Ramos Arizpe 25900, Mexico.
Sensors (Basel, Switzerland)
|June 27, 2025
Summary
This study introduces a novel system for recognizing Mexican Sign Language (LSM) by analyzing arm movements. Using pose estimation and Convolutional Neural Networks (CNNs), the system shows promising results for classifying LSM signs.
Area of Science:
- Computer Science
- Artificial Intelligence
- Linguistics
Background:
- Mexican Sign Language (LSM) recognition is an emerging field.
- Existing systems often have limitations in recognizing dynamic sign components.
Purpose of the Study:
- To develop a robust recognition system for Mexican Sign Language (LSM).
- To focus on analyzing arm movement (AM) features for sign decomposition.
Main Methods:
- Developed a proprietary dataset with participation from the deaf community and LSM experts.
- Utilized pose estimation (YOLOv8) to extract joint path data from arm movements.
- Employed Convolutional Neural Networks (CNNs) for visual pattern classification of signs.
Main Results:
- The approach successfully analyzed visual patterns in arm joint movements (wrists, shoulders, elbows).
- Pose estimation generated shapes of joint paths for CNN classification.
- Promising results were achieved in classifying sign subsets.
Conclusions:
- The proposed pose estimation-based approach demonstrates potential for building effective LSM recognition systems.
- This method can classify a wide range of signs by focusing on arm movements.
