A Survey on Posture Recognition in Bharathanatyam images

Abstract: Posture recognition is a critical element of Bharathanatyam, a classical Indian dance form, characterized by intricate footwork, graceful movements, and expressive gestures. Automating posture recognition in Bharathanatyam images is essential for preserving this cultural art form while leveraging modern technological advancements. This survey examines existing methods and technologies for posture recognition, with a specific focus on Bharathanatyam poses. It provides an extensive review of datasets, techniques, and models, including traditional machine learning approaches like SVM and HOG features, as well as advanced deep learning models such as CNNs and LSTMs. The paper highlights the challenges in dataset preparation, feature extraction, and model accuracy due to variations in dancer attributes, lighting, and pose complexity. By systematically analyzing state-of-the-art techniques and their applications in related domains, this survey aims to bridge gaps and propose future directions for improving the precision and robustness of Bharathanatyam posture recognition systems. Keywords: Bharathanatyam, posture recognition, deep learning, machine learning, feature extraction.