查看文章

Activity recognition is the process of continuously monitoring a person’s activity and movement. Human posture recognition can be utilized to assemble a self-guidance practice framework that permits individuals to learn and rehearse yoga postures accurately without help from anyone else. Using Deep learning algorithms, an approach has been taken to accurately detect and recognize various yoga poses. The chosen dataset consists of a total of 85 videos with 6 yoga postures performed by 15 participants. Initially, keypoints of the user are extracted using the Mediapipe library. A combination of convolutional neural network (CNN) and long short-term memory (LSTM) has been used for Yoga pose recognition through real-time monitored videos as a deep learning model. CNN layer is used for extraction of features from the keypoints and it is followed by LSTM that understands the occurrence of sequence of frames for predictions to be made. Then, the poses are classified as correct or incorrect. If a correct pose is identified, the system will give the user feedback through text/speech.