Yoga pose annotation and classification by using time-distributed convolutional neural network

Somashekhar S. Dhanyal, Suvarna S. Nandya


In India, people have been practicing yoga for thousands of years to improve their health and well-being on all levels. As the pace of technological development increases, this presents a great opening for computational probing across all areas of social domains. Nevertheless, it remains difficult to integrate artificial intelligence (AI) and machine learning (ML) methods to an interdisciplinary field like yoga. The proposed study aims to develop a yoga pose annotation and classification for yogasana recognition in real time. The study considers TensorFlow for better implementation of data automation, performance monitoring. TensorFlow yields better numerical computation and hat helps ML and efficiently develops the neural network. The proposed composed of time-distributed convolutional neural network (CNN) through the Softmax function. Also, a poseNet algorithm is considered to estimate the user’s real-time yoga pose. The use of a database i.e., poseTrack in the proposed method offers annotation to the evaluation of yoga pose and tracking of it. The performance analysis of the proposed yoga pose annotation and classification model suggests that it offers higher accuracy than traditional, support vector machines (SVM) and K-nearest neighbor (KNN).


Accuracy; Annotation; Classification; Convolutional neural network; Machine learning; Yoga pose

Full Text:




  • There are currently no refbacks.

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Indonesian Journal of Electrical Engineering and Computer Science (IJEECS)
p-ISSN: 2502-4752, e-ISSN: 2502-4760
This journal is published by the Institute of Advanced Engineering and Science (IAES) in collaboration with Intelektual Pustaka Media Utama (IPMU).

shopify stats IJEECS visitor statistics