<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>In this paper, firstly, in order to solve the piano playing recognition problem in the field of artificial intelligence, based on the theory of multiple intelligences, the VGG-16 deep network migration learning algorithm is applied to estimate and acquire the piano playing gesture posture. Secondly, combined with the Iterative Update Extended Kalman Filter (IUEKF) algorithm, the micro-inertial sensor fixation of the piano-playing gesture is realized, which in turn is conducive to improving the piano-playing gesture recognition accuracy. Then, we obtain real-time piano-playing gesture information through a Kinect somatosensory device, construct a piano-playing gesture recognition model based on migration learning on the basis of obtaining piano-playing gesture features, and confirm the effectiveness of the model through the experimental study of piano-playing recognition. The results show that in piano-playing gesture recognition, the recognition accuracy of this paper’s method remains above 0.9, and the application of this paper’s method can effectively improve the recognition accuracy of piano-playing gestures. On piano playing pedal action recognition, this paper’s method shows that the average F-measure scores of these two strategies are 0.924 and 0.944, respectively, which are better compared to other methods. This study provides an effective case for applying AI techniques to piano performance recognition and broadens the scope of AI applications.</p>
</abstract>

In this paper, firstly, in order to solve the piano playing recognition problem in the field of artificial intelligence, based on the theory of multiple intelligences, the VGG-16 deep network migration learning algorithm is applied to estimate and acquire the piano playing gesture posture. Secondly, combined with the Iterative Update Extended Kalman Filter (IUEKF) algorithm, the micro-inertial sensor fixation of the piano-playing gesture is realized, which in turn is conducive to improving the piano-playing gesture recognition accuracy. Then, we obtain real-time piano-playing gesture information through a Kinect somatosensory device, construct a piano-playing gesture recognition model based on migration learning on the basis of obtaining piano-playing gesture features, and confirm the effectiveness of the model through the experimental study of piano-playing recognition. The results show that in piano-playing gesture recognition, the recognition accuracy of this paper’s method remains above 0.9, and the application of this paper’s method can effectively improve the recognition accuracy of piano-playing gestures. On piano playing pedal action recognition, this paper’s method shows that the average F-measure scores of these two strategies are 0.924 and 0.944, respectively, which are better compared to other methods. This study provides an effective case for applying AI techniques to piano performance recognition and broadens the scope of AI applications.

Piano Playing Gesture Recognition Based on Multiple Intelligences Theory

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

{"article-title":"Piano Playing Gesture Recognition Based on Multiple Intelligences Theory"}

In this paper, firstly, in order to solve the piano playing recognition problem in the field of artificial intelligence, based on the theory of multiple...

Piano Playing Gesture Recognition Based on Multiple Intelligences Theory

Online veröffentlicht: 25. Nov. 2023

Seitenbereich: -

Eingereicht: 24. Jan. 2023

Akzeptiert: 05. Juni 2023

DOI: https://doi.org/10.2478/amns.2023.2.01230

Schlüsselwörter
Multiple intelligence theory, VGG-16 deep network, Transfer learning algorithm, Piano playing, Gesture recognition

© 2023 Jiahui Zhang, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Piano Playing Gesture Recognition Based on Multiple Intelligences Theory

Online veröffentlicht: 25. Nov. 2023

Seitenbereich: -

Eingereicht: 24. Jan. 2023

Akzeptiert: 05. Juni 2023

DOI: https://doi.org/10.2478/amns.2023.2.01230

SchlüsselwörterMultiple intelligence theory, VGG-16 deep network, Transfer learning algorithm, Piano playing, Gesture recognition

© 2023 Jiahui Zhang, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Schlüsselwörter
Multiple intelligence theory, VGG-16 deep network, Transfer learning algorithm, Piano playing, Gesture recognition