OpenL3 Preprocess

R2026b

Preprocess audio for OpenL3 embeddings extraction

Since R2022b

• 

OpenL3 Preprocess Block block

To add a block to a model, double-click the canvas and start typing the block name. Then, select the block from the list.

Libraries:
Audio Toolbox / Deep Learning

Description

The OpenL3 Preprocess block generates spectrograms from an audio input. You can then feed these spectrograms to an OpenL3 pretrained network or to a network that accepts the same inputs as OpenL3.

Ports

Input

Port_1 — Sound data

column vector

Sound data, specified as a one-channel signal (column vector). If Sample rate of input signal (Hz) is 48e3, there are no restrictions on the input frame length. If Sample rate of input signal (Hz) is different from 48e3, then the input frame length must be a multiple of the decimation factor of the resampling operation that the block performs. If the input frame length does not satisfy this condition, the block throws an error message with information on the decimation factor.

Data Types: single | double

Output

Port_1 — Spectrogram

matrix

Spectrogram generated from input audio, returned as a matrix whose size depends on the value of the Spectrum type parameter.

• Mel (128 bands) –– The block returns a mel spectrogram of size 128-by-199, where 128 is the number of mel bands, and 199 is the number of time hops.

• Mel (256 bands) –– The block returns a mel spectrogram of size 256-by-199, where 256 is the number of mel bands, and 199 is the number of time hops.

• Linear –– The block returns a positive one-sided spectrogram of size 257-by-197, where 257 is the FFT length and 197 is the number of time hops.

You can use this spectrogram as input to an OpenL3 block that has the same Spectrum type.

Data Types: single

Parameters

Sample rate of input signal (Hz) — Sample rate of input signal in Hz

48e3 (default) | positive scalar

Sample rate of the input signal in Hz, specified as a positive scalar.

Overlap percentage (%) — Overlap percentage between consecutive spectrograms

90 (default) | [0 100)

Specify the overlap percentage between consecutive spectrograms as a scalar in the range [0 100).

Spectrum type — Type of spectrum

Mel (128 bands) (default) | Mel (256 bands) | Linear

Type of spectrum generated from input audio, specified as Mel (128 bands), Mel (256 bands), or Linear.

Block Characteristics

Data Types	double | single
Direct Feedthrough	no
Multidimensional Signals	no
Variable-Size Signals	no
Zero-Crossing Detection	no

Programmatic Use: Add Block

To add the block programmatically, use the add_block function with one of these library block paths as the source.

"audioai/OpenL3 Preprocess"

Example: add_block("audioai/OpenL3 Preprocess","myModel/myBlock",MakeNameUnique="on")

References

[1] Cramer, Jason, et al. "Look, Listen, and Learn More: Design Choices for Deep Audio Embeddings." In ICASSP 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE, 2019, pp. 3852-56. DOI.org (Crossref), doi:/10.1109/ICASSP.2019.8682475.

Extended Capabilities

C/C++ Code Generation
Generate C and C++ code using Simulink® Coder™.

Version History

Introduced in R2022b