Identification, Classification and Modelling of Traditional African Dance Using Deep Learning Techniques.
DescriptionIdentifying patterns and their causal variables in dataset is an important component of Data Science. The field of data science has experienced enormous growth in the recent time. Many data analytics algorithms and solutions sprung up in the last decade. Data science has been useful in solving many old-age and complex problems in several domains such as healthcare, education, transportation, business and technology industries. However, the diversity of data types and representation calls for more sophisticated algorithms to make sense of image and video data. This research proposes a novel framework that applies deep learning algorithms to the field of intangible cultural preservation by applying various deep learning techniques to identify, classify and model traditional African dances from videos. Traditional African dances are important part of the African culture and heritage. Digital preservation of these dances in their multitudes and forms is a problem. Dance actions will be extracted using optical flow and pose description based on Histogram of Oriented Optical Flow (HOOF). Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), and Generative Adversarial Networks (GANs) will be used for the identification, classification and modelling process.