TFF2020 - Time-Frequency Fading

Code and data to reproduce experiments from paper Time-frequency fading algorithms based on Gabor multipliers by A. Marina Krémé, Valentin Emiya, Caroline Chaux and Bruno Torré́sani, 2020.

The sound material is available in folder 'data'.

The code is available in folders 'matlab' and 'python'. The main experiments are available in both programming languages:

• Figure 1 can be reproduced in Matlab by running file run_illustration_cuicui_eigenvalues.m
• Figure 2 can be reproduced in Matlab by running file exp_eigenval_win.m.
• Figure 3 can be reproduced in Matlab by running file rank_estimation_halko_vs_eigs_gausswin.m.
• Figure 4 can be reproduced in Python by running the specific tasks 12 and 13 from tffpy.scripts.script_exp_solve_tff.py.
• Figure 5 can be reproduced in Python by running the specific tasks 12 and 13 from tffpy.scripts.script_exp_solve_tff.py.
• Figure 6 can be reproduced in Python by running the full experiment from tffpy.scripts.script_exp_solve_tff.py.
• Table I can be reproduced in Python by running the full experiment from tffpy.scripts.script_exp_solve_tff.py.
• Table II can be reproduced in Python by running the full experiment from tffpy.scripts.script_exp_solve_tff.py.