• Extracted audio features including A-weighted coefficients, Mel-frequency cepstrum from basketball games.
• Performed crowd noise classification to identify instances of crowd cheers and boos for automatic highlight generation.
• Utilizing random grid-search hyperparameter tuning for Random Forest classifier achieved 93% training accuracy.
• Built a Support Vector Machine classifier to classify accelerometer data to gain insights about the audience’s behaviour and emotional response at a sporting event and attained 98% test accuracy.

Please sign in or create an account to give a rating or comment.

Please sign in or create an account to view the download file