Dynamic effects on reservoir computing with a Hopf oscillator

Shougat, Md Raf E Ul; Li, Xiaofu; Perkins, Edmon

doi:10.1103/physreve.105.044212

Cited by 9 publications

(4 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most nonlinear oscillator-based physical reservoir computers must use time-delayed feedback, which is cumbersome as it would require digital-to-analog and analog-to-digital converters. However, the Hopf oscillator is capable of storing enough information in its dynamic states to avoid this 24 , 25 . Moreover, the presented architecture is robust to noise because of the Hopf oscillator’s nonlinearity, which is important for real-world audio processing applications.…”

Section: Discussionmentioning

confidence: 99%

“…In this paper, we present the results of sound signal recognition using reservoir computing technology consisting of a Hopf oscillator 24 , 25 . Instead of employing computationally expensive preprocessing (e.g., Mel spectrum) commonly used in other studies 15 , 17 , 20 , 30 , we directly take the outputs from the Hopf circuit to process the normalized audio signal for machine learning recognition.…”

Section: Discussionmentioning

confidence: 99%

“…The Hopf oscillator acts as a nonlinear filter, but also some portion of the computational task is off-loaded to the Hopf physical reservoir computer. Based on our previous work, the Hopf oscillator both performs computation and stores information in its dynamic states 24 , 25 . Fundamentally, the nonlinear response of the oscillator is a type of nontraditional computing, which is unlocked through machine learning.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Hopf physical reservoir computer for reconfigurable sound recognition

Shougat

Li²,

Shao

et al. 2023

Sci Rep

Self Cite

View full text Add to dashboard Cite

The Hopf oscillator is a nonlinear oscillator that exhibits limit cycle motion. This reservoir computer utilizes the vibratory nature of the oscillator, which makes it an ideal candidate for reconfigurable sound recognition tasks. In this paper, the capabilities of the Hopf reservoir computer performing sound recognition are systematically demonstrated. This work shows that the Hopf reservoir computer can offer superior sound recognition accuracy compared to legacy approaches (e.g., a Mel spectrum + machine learning approach). More importantly, the Hopf reservoir computer operating as a sound recognition system does not require audio preprocessing and has a very simple setup while still offering a high degree of reconfigurability. These features pave the way of applying physical reservoir computing for sound recognition in low power edge devices.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Hopf physical reservoir computer for reconfigurable sound recognition

Shougat

Li²,

Shao

et al. 2023

Sci Rep

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, oscillators are capable of other types of computation as well, even without adaptive states. For instance, the classical, non-adaptive Hopf oscillator can be realized as a powerful, reconfigurable reservoir computer (Shougat et al 2021b(Shougat et al , 2022. In this reservoir computing architecture, the physics of the oscillator are utilized as a computational resource through machine learning.…”

Section: Introductionmentioning

confidence: 99%

Chaos in a Pendulum Adaptive Frequency Oscillator Circuit Experiment

Lİ

Beal

Dean

et al. 2023

Chaos Theory and Applications

View full text Add to dashboard Cite

Adaptive oscillators can learn and encode information in dynamic, plastic states. The pendulum has recently been proposed as the base oscillator of an adaptive system. In a mechanical setup, the horizontally forced pendulum adaptive frequency oscillator seeks a resonance condition by modifying the length of the pendulum's rod. This system stores the external forcing frequency when the external amplitude is small, while it can store the resonance frequency, which is affected by the nonlinearity of the pendulum, when the external amplitude is large. Furthermore, for some frequency ranges, the pendulum adaptive frequency oscillator can exhibit chaotic motion when the amplitudes are large. This adaptive oscillator could be used as a smart vibratory energy harvester device, but this chaotic region could degrade its performance by using supplementary energy to modify the rod length. The pendulum adaptive frequency oscillator’s equations of motions are discussed, and a field-programmable analog array is used as an experimental realization of this system as an electronic circuit. Bifurcation diagrams are shown for both the numerical simulations and experiments, while period-3 motion is shown for the numerical simulations. As little work has been done on the stability of adaptive oscillators, the authors believe that this work is the first demonstration of chaos in an adaptive oscillator.

show abstract