Artificial Bandwidth Extension (ABE)

Description:

This package contains scripts to perform artificial bandwidth extension (ABE), super-wide bandwidth extension (SWBE) and some useful functions

---

Copyright (C) 2018 EURECOM, France. This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/4.0/

---

Contents:

1. ABE_explicit_memory_ICASSP18

   • This package contains scripts to perform ABE using explicit memory inclusion approach presented in the following publication: P. Bachhav, M. Todisco, and N. Evans, “Exploiting explicit memory inclusion for artificial bandwidth extension,” in Proc. of IEEE Int. Conf. on Acoustics, Speech and Signal Processing (ICASSP), 2018, pp. 5459-63, Calgary, Canada.
   • Refer "Readme.txt" in folder "ABE_explicit_memory_ICASSP18" for further details

2. SWBE_LPAS_ICASSP18

   • This package contains scripts to perform SWBE using linear prediction based analysis-synthesis approach presented in the following publication: P. Bachhav, M. Todisco, and N. Evans, “Efficient super-wide bandwidth extension using linear prediction based analysis synthesis,” in Proc. of IEEE Int. Conf. on Acoustics, Speech and Signal Processing (ICASSP), 2018, pp. 5429-5433, Calgary, Canada.
   • Refer "Readme.txt" in folder "SWBE_LPAS_ICASSP18" for further details

3. ABE_SSAE_IS18

   • This package contains scripts to perform ABE with memory inclusion using Semi-supervised Stacked Auto-encoders approach presented in the following publication: P. Bachhav, M. Todisco, and N. Evans, "Artificial Bandwidth Extension with Memory Inclusion using Semi-supervised Stacked Auto-encoders", in Proc. of INTERSPEECH, pp. 1185-89, Hyderabad, India.
   • Refer "Readme.txt" in folder "ABE_SSAE_IS18" for further details

4. ABE_CVAE_ICASSP19

   • This package contains scripts to perform ABE with latent features obtained using a deep generative model based approach presented in the following publication: P. Bachhav, M. Todisco, and N. Evans, "Latent Representation Learning for Artificial Bandwidth Extension using a Conditional Variational Auto-Encoder", accepted in ICASSP 2019.
   • Refer "Readme.txt" in folder "ABE_CVAE_ICASSP19" for further details

5. utilities

   • This folder contains MATLAB scripts or functions used for ABE and SWBE tasks listed above

6. Filters

   • This folder contains impulse response of FIR filters used

7. Examples

   • This folder contains usage examples of some functions available in folder "utilities"

---

All the scripts are successfully tested on MATLAB 2018a

---

Contact information

For any query, please contact:

• Pramod Bachhav (bachhav at eurecom dot fr)
• Massimiliano Todisco (todisco at eurecom dot fr)
• Nicholas Evans (evans at eurecom dot fr)

---