Role of non-linear data processing on speech recognition task in the framework of reservoir computing.
Journal
Scientific reports
ISSN: 2045-2322
Titre abrégé: Sci Rep
Pays: England
ID NLM: 101563288
Informations de publication
Date de publication:
15 01 2020
15 01 2020
Historique:
received:
14
05
2019
accepted:
19
12
2019
entrez:
17
1
2020
pubmed:
17
1
2020
medline:
17
1
2020
Statut:
epublish
Résumé
The reservoir computing neural network architecture is widely used to test hardware systems for neuromorphic computing. One of the preferred tasks for bench-marking such devices is automatic speech recognition. This task requires acoustic transformations from sound waveforms with varying amplitudes to frequency domain maps that can be seen as feature extraction techniques. Depending on the conversion method, these transformations sometimes obscure the contribution of the neuromorphic hardware to the overall speech recognition performance. Here, we quantify and separate the contributions of the acoustic transformations and the neuromorphic hardware to the speech recognition success rate. We show that the non-linearity in the acoustic transformation plays a critical role in feature extraction. We compute the gain in word success rate provided by a reservoir computing device compared to the acoustic transformation only, and show that it is an appropriate bench-mark for comparing different hardware. Finally, we experimentally and numerically quantify the impact of the different acoustic transformations for neuromorphic hardware based on magnetic nano-oscillators.
Identifiants
pubmed: 31941917
doi: 10.1038/s41598-019-56991-x
pii: 10.1038/s41598-019-56991-x
pmc: PMC6962256
doi:
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
328Subventions
Organisme : EC | EU Framework Programme for Research and Innovation H2020 | H2020 Priority Excellent Science | H2020 European Research Council (H2020 Excellent Science - European Research Council)
ID : bioSPINspired 682955
Pays : International
Organisme : EC | EU Framework Programme for Research and Innovation H2020 | H2020 Priority Excellent Science | H2020 European Research Council (H2020 Excellent Science - European Research Council)
ID : bioSPINspired 682955
Pays : International
Organisme : EC | EU Framework Programme for Research and Innovation H2020 | H2020 Priority Excellent Science | H2020 European Research Council (H2020 Excellent Science - European Research Council)
ID : bioSPINspired 682955
Pays : International
Références
LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. Nature 521, 436 (2015).
doi: 10.1038/nature14539
Mead, C. Neuromorphic electronic systems. Proc. IEEE 78, 1629–1636, https://doi.org/10.1109/5.58356 (1990).
doi: 10.1109/5.58356
Mead, C. & Ismail, M. (eds) Analog VLSI implementation of neural systems, vol. 80 (Springer Science & Business Media, 2012).
Maass, W., Natschläger, T. & Markram, H. Real-time computing without stable states: A new framework for neural computation based on perturbations. Neural Comput. 14, 2531–2560, https://doi.org/10.1162/089976602760407955 (2002).
doi: 10.1162/089976602760407955
pubmed: 12433288
Jaeger, H. & Haas, H. Harnessing nonlinearity: Predicting chaotic systems and saving energy in wireless communication. Science 304, 78–80, https://doi.org/10.1126/science.1091277 (2004).
doi: 10.1126/science.1091277
pubmed: 15064413
Appeltant, L. et al. Information processing using a single dynamical node as complex system. Nat. Commun. 2, 468 (2011).
doi: 10.1038/ncomms1476
Paquot, Y. et al. Optoelectronic reservoir computing. Sci. Reports 2, 287 (2012).
doi: 10.1038/srep00287
Torrejon, J. et al. Neuromorphic computing with nanoscale spintronic oscillators. Nature 547, 428 (2017).
doi: 10.1038/nature23011
Davis, S. & Mermelstein, P. Comparison of parametric representations for monosyllabic word recognition in continuously spoken sentences. IEEE Transactions on Acoust. Speech, Signal Process. 28, 357–366, https://doi.org/10.1109/TASSP.1980.1163420 (1980).
doi: 10.1109/TASSP.1980.1163420
Verstraeten, D., Schrauwen, B., Stroobandt, D. & Van Campenhout, J. Isolated word recognition with the liquid state machine: a case study. Inf. Process. Lett. 95, 521–528, https://doi.org/10.1016/j.ipl.2005.05.019 , Applications of Spiking Neural Networks (2005).
doi: 10.1016/j.ipl.2005.05.019
Hinton, G. et al. Deep neural networks for acoustic modeling in speech recognition. IEEE Signal Process. Mag. 29, 82–97 (2012).
doi: 10.1109/MSP.2012.2205597
Lyon, R. A computational model of filtering, detection, and compression in the cochlea. In ICASSP ’82. IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 7, 1282–1285, https://doi.org/10.1109/ICASSP.1982.1171644 (1982).
Slaney, M. Lyon’s Cochlear Model. Apple Computer, Inc., Cupertino, CA, apple computer technical report#13 edn. (1988).
Brunner, D., Soriano, M. C., Mirasso, C. R. & Fischer, I. Parallel photonic information processing at gigabyte per second data rates using transient states. Nat. Commun. 4, 1364 (2013).
doi: 10.1038/ncomms2368
Vandoorne, K. et al. Experimental demonstration of reservoir computing on a silicon photonics chip. Nat. Commun. 5, 3541 (2014).
doi: 10.1038/ncomms4541
Riou, M. et al. Neuromorphic computing through time-multiplexing with a spin-torque nano-oscillator. Electron Devices Meet. (IEDM), 2017 IEEE Int. 36.3.1–36.3.4, https://doi.org/10.1109/IEDM.2017.8268505 (2017).
Larger, L. et al. Photonic information processing beyond turing: an optoelectronic implementation of reservoir computing. Opt. Express 20, 3241–3249, https://doi.org/10.1364/OE.20.003241 (2012).
doi: 10.1364/OE.20.003241
pubmed: 22330562
Dejonckheere, A. et al. All-optical reservoir computer based on saturation of absorption. Opt. Express 22, 10868–10881, https://doi.org/10.1364/OE.22.010868 (2014).
doi: 10.1364/OE.22.010868
pubmed: 24921786
Vinckier, Q. et al. High-performance photonic reservoir computer based on a coherently driven passive cavity. Optica 2, 438–446, https://doi.org/10.1364/OPTICA.2.000438 (2015).
doi: 10.1364/OPTICA.2.000438
Brunner, D. et al. Tutorial: Photonic neural networks in delay systems. J. Appl. Phys. 124, 152004, https://doi.org/10.1063/1.5042342 (2018).
doi: 10.1063/1.5042342
Penkovsky, B., Larger, L. & Brunner, D. Efficient design of hardware-enabled reservoir computing in fpgas. J. Appl. Phys. 124, 162101, https://doi.org/10.1063/1.5039826 (2018).
doi: 10.1063/1.5039826
Duport, F., Schneider, B., Smerieri, A., Haelterman, M. & Massar, S. All-optical reservoir computing. Opt. Express 20, 22783–22795, https://doi.org/10.1364/OE.20.022783 (2012).
doi: 10.1364/OE.20.022783
pubmed: 23037429
Martinenghi, R., Rybalko, S., Jacquot, M., Chembo, Y. K. & Larger, L. Photonic nonlinear transient computing with multiple-delay wavelength dynamics. Phys. Rev. Lett. 108, 244101, https://doi.org/10.1103/PhysRevLett.108.244101 (2012).
doi: 10.1103/PhysRevLett.108.244101
pubmed: 23004274
Romain Martinenghi, M. J. Y. K. C. L. L., Antonio Baylon Fuentes. Towards optoelectronic architectures for integrated neuromorphic computers. Proc. SPIE OPTO 8989, https://doi.org/10.1117/12.2038347 (2014).
Larger, L. et al. High-speed photonic reservoir computing using a time-delay-based architecture: Million words per second classification. Phys. Rev. X 7, 011015, https://doi.org/10.1103/PhysRevX.7.011015 (2017).
doi: 10.1103/PhysRevX.7.011015
Lukoševičius, M. & Jaeger, H. Reservoir computing approaches to recurrent neural network training. Comput. Sci. Rev. 3, 127–149, https://doi.org/10.1016/j.cosrev.2009.03.005 (2009).
doi: 10.1016/j.cosrev.2009.03.005
van der Maaten, L. & Hinton, G. Visualizing data using t-sne. J. machine learning research 9, 2579–2605 (2008).
Slavin, A. N. & Tiberkevich, V. S. Nonlinear auto-oscillator theory of microwave generation by spin-polarized current. IEEE Transactions on Magn. 45, 1875–1918, https://doi.org/10.1109/TMAG.2008.2009935 (2009).
doi: 10.1109/TMAG.2008.2009935
Grimaldi, E. et al. Response to noise of a vortex based spin transfer nano-oscillator. Phys. Rev. B 89, 104404, https://doi.org/10.1103/PhysRevB.89.104404 (2014).
doi: 10.1103/PhysRevB.89.104404
Romera, M. et al. Vowel recognition with four coupled spin-torque nano-oscillators. Nature 563, 230–234, https://doi.org/10.1038/s41586-018-0632-y (2018).
doi: 10.1038/s41586-018-0632-y
pubmed: 30374193
Tsunegi, S. et al. High emission power and q factor in spin torque vortex oscillator consisting of feb free layer. Appl. Phys. Express 7, 063009 (2014).
doi: 10.7567/APEX.7.063009
Hirsch, H. & Pearce, D. The aurora experimental framework for the performance evaluation of speech recognition systems under noisy conditions. In Proceedings of ISCA ITRW ASR2000 on Automatic Speech recognition: Challenges for the next Millennium (Paris, France, 2000).
ELRA catalogue ( http://catalog.elra.info ), AURORA Project Database, v2.0, ISLRN: 977-457-139-304-2, ELRA ID: AURORA/CD0002.
Leonard, R. & Doddington, G. TIDigits. (Linguistic Data Consortium, Philadelphia, 1993).