ADAPTIVE FILTERING AND MACHINE LEARNING METHODS IN NOISE SUPPRESSION SYSTEMS, IMPLEMENTED ON THE SoC
DOI:
https://doi.org/10.15588/1607-3274-2024-4-16Keywords:
embedded systems, system-on-a-chip, FPGA, adaptive filtering, digital signal processing algorithms, noise suppression algorithms, audio signals, machine learning, neural networksAbstract
Context. Modern video conferencing systems work in different noise environments, so preservation of speech clarity and provision of quick adaptation to changes in this environment are relevant tasks. During the development of embedded systems, finding a balance between resource consumption, performance, and signal quality obtained after noise suppression is necessary. Systems on a chip allow us to use the power of both processor cores available on the hardware platform and FPGAs to perform complex calculations, which contributes to increasing the speed or reducing the load on the central SoC cores.
Objective. To conduct a comparative analysis of the noise suppression quality in audio signals by an adaptive filtering algorithm and a filtering algorithm using machine learning based on the RNNoise neural network in noise suppression devices on the technological platform SoC.
Method. Evaluation using objective metrics and spectrogram analysis using the Librosa library in Python. Neural network training and model design are performed on the basis of Python and Torch tools. The Vitis IDE package was used for the neural network implementation on the platform SoC.
Results. The analysis of two noise suppression methods using the adaptive Wiener filter and the RNNoise neural network was performed. In the considered scenarios, it was determined that the neural network shows better noise suppression results according to the analysis of spectrograms and objective metrics.
Conclusions. A comparative analysis of the effectiveness of noise suppression algorithms based on adaptive filters and a neural network was performed for scenarios with different noise environments. The results of objective SIGMOS metrics were obtained to evaluate the quality of the received audio signal. In addition, the possibility of running the RNNoise neural network on the technological platform SoC ZYNQ 7000 was verified.
References
Lahti S., Rintala M., Hamalainen T. D. Leveraging Modern C++ in High-level Synthesis, IEEE Transactions on ComputerAided Design of Integrated Circuits and Systems, 2023, Vol. 42, № 4, pp. 1123–1132. DOI: 10.1109/TCAD.2022.3193646.
Monson J., Wirthlin M., Hutchings B. L. Optimization techniques for a high level synthesis implementation of the Sobel filter, International Conference on reconfigurable computing and FPGAs (ReConFig’13). Cancun, Mexico, 9–11 December 2013, pp. 1–6. DOI: 10.1109/ReConFig.2013.6732315.
Plugariu O., Petrica L., Pirea R., Hobincu R. Hadoop ZedBoard cluster with GZIP compression FPGA acceleration, 11th International Conference on Electronics, Computers and Artificial Intelligence (ECAI’19). Pitesti, Romania, 27–29 June 2019, pp. 1–5. DOI: 10.1109/ ecai46879.2019.9042006.
Rong X., Sun T., Zhang X., Hu Y., Zhu C. , Lu J. GTCRN: A speech enhancement model requiring ultralow computational resources, IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP’2024). Seoul, Korea, 14–19 April 2024, pp. 971–975. DOI: 10.1109/icassp48485.2024.10448310.
Leiva L., Torrents-Barrena J., Vazquez M. FPGA-based accelerator for AI-toolbox reinforcement learning library, IEEE Embedded Systems Letters, 2023, Vol. 15, № 2, pp. 113–116. DOI: 10.1109/les.2022.3218168.
Fan H., Wang H., Che K., Wu Z. Design of FPGA deep neural network accelerator based on high-level synthesis, 5th International Academic Exchange Conference on Science and Technology Innovation (IAECST’23). Guangzhou, China, 8–10 December 2023, pp. 163–166. DOI: 10.1109/iaecst60924.2023.10502749.
Chang Y.-W., Huang C.-C., Hwang Y.-T. A face mask detection system based on high level synthesis and hardware software codesign, IET International Conference on Engineering Technologies and Applications (IET-ICETA’22). Changhua, Taiwan, 14–16 October 2022, pp. 1–2. DOI: 10.1109/ieticeta56553.2022. 9971488.
Riazati M., Daneshtalab M., Sjödin M. , Lisper B. DeepFlexiHLS: Deep neural network flexible high-level synthesis directive generator, IEEE Nordic Circuits and Systems Conference (NorCAS’22). Oslo, Norway, 25–26 October 2022, pp. 1–6. DOI: 10.1109/norcas 57515.2022.9934617.
Doumanidis C. C. , Anagnostou C. , Arvaniti E.-S., Papadopoulou A. RNNoise-Ex: hybrid speech enhancement system based on rnn and spectral features [Electronic resource], 2021, pp. 1–5. Access mode: https://arxiv.org/pdf/2105.11813. DOI: 10.48550/arXiv. 2105.11813
Cheng B., Zhang G., Tao X. , Wang S., Wu N., Chen M. An improved real-time noise suppression method based on RNN and long-term speech information, 3rd international symposium on automation, information and computing (ICSPCC’22). Beijing, China, China, 9–11 December 2022, pp. 476–481. DOI: 10.1109/ICASSP40776.2020.9054597.
Reddy C.K.A., Beyrami E., Pool J., Cutler R., Srinivasan S., Gehrke J. A scalable noisy speech dataset and online subjective test framework, International Conference “Interspeech 2019”. Graz, Austria, 15–19 September 2019. pp. 1816–1820. DOI: 10.21437/interspeech. 2019-3087.
Chen B. , Zhou Y., Ma Y., Liu H. A new real-time noise suppression algorithm for far-field speech communication based on recurrent neural network, IEEE International Conference on Signal Processing, Communications and Computing (ICSPCC’22). Xi’an, China, 17–19 August 2021, pp. 1–5. DOI: 10.1109/icspcc52875. 2021.9564530.
Zenodo: Python library for audio and music analysis Librosa [Electronic resource], 2024. Access mode: https://zenodo.org/records/11192913.
Scheibler R., Bezzam E., Dokmanić I. Pyroomacoustics: A Python package for audio room simulation and array processing
algorithms, IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP’18), Calgary. Canada, 15–20 April 2018, pp. 351–355. DOI: doi.org/10.1109/icassp.2018.8461310.
Upadhyay N., Jaiswal R. K. Single channel speech enhancement: using wiener filtering with recursive noise estimation, Proceeding of the Seventh International Conference on Intelligent Human Computer Interaction, 2016, Vol. 84, pp. 22–30. DOI: 10.1016/j.procs.2016.04.061.
Shkil A., Rahlis D., Filippenko I. , Kornijenko V., Rozhnova T. Automated design of embedded digital signal processing systems on SOC platform, Innovative technologies and scientific solutions for industries, 2024, No. 1 (27), pp. 192– 203. DOI: https://doi.org/10.30837/ITSSI.2024.27.192
Naderi B., Cutler R. Subjective evaluation of noise suppression algorithms in crowdsourcing, International Conference “Interspeech 2021”. Brno, Czechia, 30 August –3 September 2021, pp. 2132–2136. DOI: 10.21437/interspeech.2021-343.
Catalin R. N., Saabas A., Cutler R., Naderi B. , Braun S., Branets S. Speech signal improvement challenge [Electronic resource], IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP’24), 2024. Access mode: https://arxiv.org/pdf/2401.14444. DOI: 10.48550/arXiv.2401.14444.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2024 A. S. Shkil, O. I. Filippenko, D. Y. Rakhlis, I. V. Filippenko, A. V. Parkhomenko, V. R. Korniienko
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Creative Commons Licensing Notifications in the Copyright Notices
The journal allows the authors to hold the copyright without restrictions and to retain publishing rights without restrictions.
The journal allows readers to read, download, copy, distribute, print, search, or link to the full texts of its articles.
The journal allows to reuse and remixing of its content, in accordance with a Creative Commons license СС BY -SA.
Authors who publish with this journal agree to the following terms:
-
Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License CC BY-SA that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
-
Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
-
Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.
