Gave an invited talk at Academia Sinica on “Revamping latent variable analysis for speech and audio processing.”