MVICAD2: Multi-View Independent Component Analysis with Delays and Dilations

Apple researchers have developed MVICAD2, a machine learning method designed to analyze complex, multi-view datasets that incorporate time delays and variations in signal intensity. This technique is intended to help align disparate data sources, such as neural recordings from multiple subjects, by isolating common underlying signals while accounting for individual differences and sensor biases.