Oscillator-Based Processing Unit for Formant Recognition
Tamás Rudner-Halász, Wolfgang Porod, Gyorgy Csaba
Abstract
Oscillatory neural networks have been successfully applied to a number of computing problems, such as associative memories and computationally hard optimization tasks. In this paper, we show how to use oscillators to process time-dependent waveforms with minimal or no preprocessing. Since preprocessing and first-layer processing are often the most power-hungry steps in neural networks, our findings may open new doors to simple and power-efficient edge-AI devices.
Create a lesson
Related papers
A Scaling Framework for Mechanical Memristance: Dimensionless Metrics and Material Design Maps
Abdulla Alhembar, Fabrizio Scarpa, Chrystel D. L. Remillat et al.
Fast, optimal readout of a tethered optomechanical pressure sensor
Stephen R. Chen, Yiliang Bao, John R. Lawall et al.
Pushing the Dose Limit of Atomic-Resolution Imaging: A 4D-STEM case study of NaCl
Tamazouzt Chennit, Arno Annys, Songge Li et al.
A Review of ERA5 Equatorial Wind Data and Impact on a Space Elevator
Blaise Gassend, Jonathan Gassend
Shape- and Cation-Engineered Ferrite Nanoparticles for Enhanced Theranostic Performance in Ovarian Cancer Tumor-Mimicking Phantom
Bahareh Rezaei, Md Shahriar, Shahriar Mostufa et al.
Bimorph Lithium Niobate Thickness-Shear Overtone Film Bulk Acoustic Resonator
Ziqian Yao, Ian Anderson, Tzu-Hsuan Hsu et al.