Auditory Analysis of Tabla solo performances A study of Gharana specific characteristics

Tabla is a pounding instrument in Hindustani classical music tradition. Tabla learning and presentation in the Indian landmass is based on stylistic schools called Gharana. Each Gharana is attributed by its main style of playing skills, act of Tabla beats, repertoire, compositions and improvisations. Recognizing the Gharana from a Tabla presentation is mainly helpful to set apparat the performance. The paper address the work of pre-programmed Gharana identified from solo Tabla recordings. I motivate the challenge and show various aspects and provocations in the task. I recognize an easy and diverse collection of over 16 hours of Tabla Sola recordings for the task. I suggest an approach using deep learning models that use an amalgam of convolutional neural networks(CNN) and long short term memory(LSTM) networks. The CNNs are used to draw out Gharana unfair features from the raw audio data. The LSTM networks are worth to classify the Gharana by using the sequence of draw out features from CNNs. Our demonstrations on Gharana recognition include different length of audio data and contrast between various aspect of the task. An expansion demonstrates a good result with highest recognition accuracy of 92% of Hindustani music as it keeps track of rhythm. It is not used only an accompaniment but also used in solo performances. Tabla solo is complex and elaborate, with a variety of pre-composed forms for uplifting further elaborations based on the player’s stylistic schools called