Computer Vision: 3D based reconstruction of foregrounds in the video onther test with manual calibration of 4 cameras.

Identify speaker from given speech signal using MFCC features and Gaussian Mixture Models

Unsupervised Learning - Analyze feature relevance, feature distribution, scatter matrix, heatmap, feature scaling, detect and remove outliers, PCA, dimensionality reduction, Gaussian Mixture Model clustering, data recovery by inverse transformations

Gravitational-wave population inference allowing narrow population features with minimal edge bias