2020/07/08 by Dawid Przyczyna, Maria Szacilowska, Przyczyna, Dawid +9
Computer Science · Engineering · Neuroscience · #Advanced Memory and Neural Computing #Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Neural Networks and Applications #Neural Networks and Reservoir Computing #Neural dynamics and brain function #Sound (cs.SD) #cs.SD #eess.AS #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2007.04360
arxiv created 2020/07/08 · openalex publication_date 2020/07/08 · arxiv updated 2020/07/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Reservoir computing is an emerging, but very successful approach towards processing and classification of various signals. It can be described as a model of a transient computation, where influence of input changes internal dynamics of chosen computational reservoir. Trajectory of these changes represents computation performed by the system. The selection of a suitable computational substrate capable of non-linear response and rich internal dynamics ensures the implementation of simple readout protocols. Signal evolution based on the rich dynamics of the reservoir layer helps to emphasize differences between given signals thus enabling their easier classification. Here we present a simple reservoir computing system (single node echo-state machine) implemented on Multisim platform as a tool for classification of musical intervals according to their consonant or dissonant character. The result of this classification closely resembled sensory dissonance curve, with some significant differences. A deeper analysis of the received signals indicates the geometric relationships between the consonant and dissonant intervals, enabling their classification.