Hand Gesture Classification Technique Based on Acceleration and Ultrasound Data Fusion for Human Movement Recognition and Device Control 


Vol. 45,  No. 12, pp. 2140-2149, Dec.  2020
10.7840/kics.2020.45.12.2140


PDF
  Abstract

As IT technology grows, wearable devices and Internet of Things devices are increasing, and they are becoming more common for people. As devices get smaller and smaller in size for portability, it is becoming increasingly difficult to control using general buttons or touches. To solve this problem, studies are being conducted on a new interface that controls a device by recognizing human motion. In this paper, we present a method of classifying hand gestures with only the microphone and speaker in the smart phone, and accelerometer built into the smart watch without using an additional sensor. We propose a method to increase the accuracy of hand gesture classification by fusion of acceleration data and ultrasound data. According to the fusion operation, it is classified into Max, Add, and Concat models. These fusion models improved classification accuracy compared to models that train only one part of the data. The Concat model, which is the best performing model showed 90.0% classification in 10 patterns, this classification accuracy is improved by 5.8% compared to the model trained using only acceleration data and 16.4% compared to the model using only ultrasound data.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Cite this article

[IEEE Style]

J. Cheon and S. Choi, "Hand Gesture Classification Technique Based on Acceleration and Ultrasound Data Fusion for Human Movement Recognition and Device Control," The Journal of Korean Institute of Communications and Information Sciences, vol. 45, no. 12, pp. 2140-2149, 2020. DOI: 10.7840/kics.2020.45.12.2140.

[ACM Style]

Jinwon Cheon and Sunwoong Choi. 2020. Hand Gesture Classification Technique Based on Acceleration and Ultrasound Data Fusion for Human Movement Recognition and Device Control. The Journal of Korean Institute of Communications and Information Sciences, 45, 12, (2020), 2140-2149. DOI: 10.7840/kics.2020.45.12.2140.

[KICS Style]

Jinwon Cheon and Sunwoong Choi, "Hand Gesture Classification Technique Based on Acceleration and Ultrasound Data Fusion for Human Movement Recognition and Device Control," The Journal of Korean Institute of Communications and Information Sciences, vol. 45, no. 12, pp. 2140-2149, 12. 2020. (https://doi.org/10.7840/kics.2020.45.12.2140)

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880