Related Experiment Video
Updated: Dec 21, 2025
08:15
Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
Published on: March 28, 2025
1.1K
HGM-4: A new multi-cameras dataset for hand gesture recognition
1Ho Chi Minh City Open University, Vietnam.
Data in Brief
|May 22, 2020
Summary
This study introduces the HGM-4 dataset for hand gesture recognition, featuring 4,160 images of 26 gestures. This dataset provides a benchmark for developing advanced gesture recognition systems.
Area of Science:
- Computer Vision
- Human-Computer Interaction
Background:
- Gesture recognition technology is rapidly advancing, driven by applications in gaming, robotics, and smart homes.
- Hand-gesture recognition is a key area, enabling device control through natural movements.
Purpose of the Study:
- To introduce and describe the HGM-4 dataset for hand gesture recognition.
- To establish a benchmark framework for evaluating hand gesture recognition algorithms.
Main Methods:
- The HGM-4 dataset comprises 4,160 color images (1280x700 pixels) of 26 distinct hand gestures.
- Images were captured using four cameras from various positions to ensure diverse data.
- The dataset is divided into training and testing sets for standardized evaluation.
Main Results:
- The HGM-4 dataset provides a comprehensive resource for hand gesture recognition research.
- The defined training and testing sets facilitate direct comparison of experimental results.
Conclusions:
- The HGM-4 dataset serves as a valuable benchmark for advancing hand gesture recognition technology.
- This resource supports the development of more sophisticated gesture-controlled applications.
