A Feasibility Test on Preventing PRMDs Based on Deep Learning

Abstract

This study proposes a method to reduce the playing-related musculoskeletal disorders (PRMDs) that often occur among pianists. Specifically, we propose a feasibility test that evaluates several state-of-the-art deep learning algorithms to prevent injuries of pianist. For this, we propose (1) a C3P dataset including various piano playing postures and show (2) the application of four learning algorithms, which demonstrated their superiority in video classification, to the proposed C3P datasets. To our knowledge, this is the first study that attempted to apply the deep learning paradigm to reduce the PRMDs in pianist. The experimental results demonstrated that the classification accuracy is 80% on average, indicating that the proposed hypothesis about the effectiveness of the deep learning algorithms to prevent injuries of pianist is true.

Audio-Visual Tensor Fusion Network for Piano Player Posture Classification

“…We set the input resolution to 80 × 120 and the stride size to 5. From the graph, we can observe that as the kernel size increased, the F1 score also gradually increased, exhibiting the highest score at (6,6) and (7,7). On the other hand, Figure 9b shows the changes in the F1 score for posture classification with increasing CNN kernel size, which is not square.…”

“…Third, this study demonstrates the superiority of the proposed AV-TFN method through comparisons of the performances of the visual network (VN) [7], audio network (AN), and audio-visual network (AVN) with concatenation (AVN-Concat) [8] and attention (AVN-Atten) [9] techniques. The experiment results show that AV-TFN significantly improves F1 score compared with AN, VN, AVN-Concat, and AVN-Atten methods, while also achieving speeds similar to that of the fast VN method.…”

“…The authors of [13] used principal component analysis (PCA), a statistical procedure, to analyze pianists' posture and classify them into various levels. The authors of [7] performed a feasibility test that recommends a suitable deep learning algorithm to determine the correct posture for a pianist. Though this is not a study on posture classification, [14] proposed a Long Short-Term Memory (LSTM)-based model for predicting playing posture from sound while emphasizing the relationship between posture and sound.…”

“…Thus, we have adopted several methods from different domains. Specifically, we compared the proposed AV-TFN with VN [7] and AN that utilize only visual and audio networks, respectively. Besides, we compared the proposed method with AVN_Concat [8] and AVN_Atten [9] that use audio-visual information with various data representation techniques.…”

Forearm muscles activity of harp players

“…Training procedures are also explored to improve the musician's force, and precision (Gorniak et al, 2019;Muramatsu et al, 2022). To obtain quantitative insights, a few biomechanical studies have been recently published (Blanco-Piñeiro et al, 2017;Goubault et al, 2021;Metcalf et al, 2014;Park et al, 2019). Kinematics has mostly been interesting to address playing posture and ancillary gestures (or accompanist gestures) (Wanderley et al, 2005).…”