Fast local sparsity

The fast local sparsity (FLS) is a local information method that employs a local Hoyer sparsity measure to compute combination weights. Compared to other local information methods, the FLS is computationally efficient, as it does not require sorting vectors, and has been shown to achieve great performance in time-frequency resolution.

Further reading

• Package release article: To be added.

• FLS reference paper: M. do V. M. da Costa and L. W. P. Biscainho, “The fast local sparsity method: A low-cost combination of time-frequency representations based on the hoyer sparsity,” Journal of the Audio Engineering Society, vol. 70, no. 9, pp. 698–707, Sep. 2022.

Calling signature

ctfr.methods.fls(signal, sr, *, <shared parameters>, lk, lm, gamma)

ctfr.methods.fls_from_specs(specs, *, <shared parameters>, lk, lm, gamma)

Note

As with all combination methods, you can also use ctfr.ctfr() or ctfr.ctfr_from_specs().

See ctfr.ctfr() and ctfr.ctfr_from_specs() for more details on the shared parameters for computing CTFRs with this package. The parameters specific to this method (passed as keyword arguments) are described below.

Parameters

lk (int > 0, odd, optional)

Width in frequency bins of the analysis window used in the local sparsity computation. Defaults to 21.

lm (int > 0, odd, optional)

Width in time frames of the analysis window used in the local sparsity computation. Defaults to 11.

gamma (float >= 0, optional)

Factor used in the computation of combination weights. Defaults to 20.