The research demonstrates that a Vision Transformer adapted for audio data outperforms both a custom CNN and traditional machine learning models in classifying music genres using mel spectrograms from the GTZAN dataset.
Developed MoodMeter.AI, an advanced tool for remote work communication, employing real-time emotion detection through facial expression and voice analysis to enhance understanding in online meetings
Real-time semantic scene understanding is essential for autonomous vehicles, especially in off-road driving where it deals with uncertainties like uneven terrain and hidden obstacles, and this project aims to enhance off-road semantic segmentation to improve route visualization and navigation.
This package provides the probability distribution classes.