Software resources Software for feature extraction Librosa is a Python package that includes acoustic feature calculation MIRtoolbox is a MATLAB toolbox for music information retrieval, but has functions for feature extraction that can also be used in speech analysis OpenSMILE is a C++ based feature extraction library Speech recognition tools Kaldi is…
A test of ICA algorithms with instantaneous mixtures This code is a standard test of ICA algorithms. An instantaneous mixture of J signals is generated where the mixing matrix is random. The mixture is separated by selected methods and the standard criterion Signal-to-Interference Ratio (SIR) is evaluated on each separated signal. This test can…
NMF software openBliSSART is a C++ library for NMF-based blind source separation FASST is a toolbox for source separation which includes variants of NMF. There are C++ and MATLAB versions, as well as a third-party Python version NMF algorithms for supervised separation of known sources
Speech Enhancement in Non-Stationary Noise Original Minimum statistics spectral subtraction [1, 2] LSTM denoising [3] Sparse NMF [4] Exemplar-based sparse NMF [5] Multi-condition example 1: TUM NAVIC corpus, English, City noise (bicycle) @ 5 dB(A) Multi-condition example 2: TUM NAVIC corpus, English, Music noise @ 5 dB(A) Application to real…
Exercises: Exercise 1: GCC-PHAT & Acoustic Maps Given a speech signal of a static human source recorded in a real multi-channel acquisition set up, compute the GCC-PHAT focusing on: temporal evolution of the GCC-PHAT due to speech sparsity; behaviour of GCC-PHAT at different microphone pairs. Using the computed GCC-PHAT, derive…
