-
Hardware Accelerator and Neural Network Co-Optimization for Ultra-Low-Power Audio Processing Devices
Authors:
Christoph Gerum,
Adrian Frischknecht,
Tobias Hald,
Paul Palomero Bernardo,
Konstantin Lübeck,
Oliver Bringmann
Abstract:
The increasing spread of artificial neural networks does not stop at ultralow-power edge devices. However, these very often have high computational demand and require specialized hardware accelerators to ensure the design meets power and performance constraints. The manual optimization of neural networks along with the corresponding hardware accelerators can be very challenging. This paper present…
▽ More
The increasing spread of artificial neural networks does not stop at ultralow-power edge devices. However, these very often have high computational demand and require specialized hardware accelerators to ensure the design meets power and performance constraints. The manual optimization of neural networks along with the corresponding hardware accelerators can be very challenging. This paper presents HANNAH (Hardware Accelerator and Neural Network seArcH), a framework for automated and combined hardware/software co-design of deep neural networks and hardware accelerators for resource and power-constrained edge devices. The optimization approach uses an evolution-based search algorithm, a neural network template technique, and analytical KPI models for the configurable UltraTrail hardware accelerator template to find an optimized neural network and accelerator configuration. We demonstrate that HANNAH can find suitable neural networks with minimized power consumption and high accuracy for different audio classification tasks such as single-class wake word detection, multi-class keyword detection, and voice activity detection, which are superior to the related work.
△ Less
Submitted 29 September, 2022; v1 submitted 8 September, 2022;
originally announced September 2022.
-
Behavior of Keyword Spotting Networks Under Noisy Conditions
Authors:
Anwesh Mohanty,
Adrian Frischknecht,
Christoph Gerum,
Oliver Bringmann
Abstract:
Keyword spotting (KWS) is becoming a ubiquitous need with the advancement in artificial intelligence and smart devices. Recent work in this field have focused on several different architectures to achieve good results on datasets with low to moderate noise. However, the performance of these models deteriorates under high noise conditions as shown by our experiments. In our paper, we present an ext…
▽ More
Keyword spotting (KWS) is becoming a ubiquitous need with the advancement in artificial intelligence and smart devices. Recent work in this field have focused on several different architectures to achieve good results on datasets with low to moderate noise. However, the performance of these models deteriorates under high noise conditions as shown by our experiments. In our paper, we present an extensive comparison between state-of-the-art KWS networks under various noisy conditions. We also suggest adaptive batch normalization as a technique to improve the performance of the networks when the noise files are unknown during the training phase. The results of such high noise characterization enable future work in develo** models that perform better in the aforementioned conditions.
△ Less
Submitted 15 September, 2021;
originally announced September 2021.
-
The origin of mechanical enhancement in polymer nanoparticle composites with ultra-high nanoparticle loading
Authors:
Emily Y. Lin,
Amalie L. Frischknecht,
Robert A. Riggleman
Abstract:
Polymer nanoparticle composites (PNC) with ultra high loading of nanoparticles (> 50%) have been shown to exhibit markedly improved strength, stiffness, and toughness simultaneously compared to the neat systems of their components. Recent experimental studies on the effect of polymer fill fraction in these highly loaded PNCs reveal that even at low polymer fill fractions, hardness and modulus incr…
▽ More
Polymer nanoparticle composites (PNC) with ultra high loading of nanoparticles (> 50%) have been shown to exhibit markedly improved strength, stiffness, and toughness simultaneously compared to the neat systems of their components. Recent experimental studies on the effect of polymer fill fraction in these highly loaded PNCs reveal that even at low polymer fill fractions, hardness and modulus increase significantly. In this work, we aim to understand the origin of these performance enhancements by examining the dynamics of both polymer and nanoparticles (NP) under tensile deformation. We perform molecular dynamics (MD) simulations of coarse-grained, glassy polymer in random-close-packed nanoparticle packings with a varying polymer fill fraction. We characterize the mechanical properties of the PNC systems, compare the NP rearrangement behavior, and study the polymer segmental and chain-level dynamics during deformation below the polymer glass transition. Our simulation results confirm the experimentally-observed increase in modulus at low polymer fill fractions, and we provide evidence that the source of mechanical enhancement is the polymer bridging effect.
△ Less
Submitted 18 December, 2019;
originally announced December 2019.
-
Electrostatics of Colloidal Particles Confined in Nanochannels: Role of Double-Layer Interactions and Ion-Ion Correlations
Authors:
Inderbir S. Sidhu,
Amalie L. Frischknecht,
Paul J. Atzberger
Abstract:
We perform computational investigations of electrolyte-mediated interactions of charged colloidal particles confined within nanochannels. We investigate the role of discrete ion effects, valence, and electrolyte strength on colloid-wall interactions. We find for some of the multivalent charge regimes that the like-charged colloids and walls can have attractive interactions. We study in detail thes…
▽ More
We perform computational investigations of electrolyte-mediated interactions of charged colloidal particles confined within nanochannels. We investigate the role of discrete ion effects, valence, and electrolyte strength on colloid-wall interactions. We find for some of the multivalent charge regimes that the like-charged colloids and walls can have attractive interactions. We study in detail these interactions and the free energy profile for the colloid-wall separation. We find there are energy barriers and energy minima giving preferred colloid locations in the channel near the center and at a distance near to but separated from the channel walls. We characterize contributions from surface overcharging, condensed layers, and overlap of ion double-layers. We perform our investigations using Coarse-Grained Brownian Dynamics simulations (BD), classical Density Functional Theory (cDFT), and mean-field Poisson-Boltzmann Theory (PB). We discuss the implications of our results for phenomena in nanoscale devices.
△ Less
Submitted 28 March, 2018; v1 submitted 5 August, 2017;
originally announced August 2017.
-
Surface-induced First Order Transition in Athermal Polymer/Nanoparticle Blends
Authors:
E. S. McGarrity,
A. L. Frischknecht,
L. J. D. Frink,
M. E. Mackay
Abstract:
We investigate the phase behavior of athermal polymer/nanoparticle blends near a hard substrate. We apply the density functional theory of Tripathi and Chapman to these blends. We find a first order phase transition where the nanoparticles expel the polymer from the surface to form a monolayer. The transition density depends on the polymer length and the system bulk density. The effect is due to…
▽ More
We investigate the phase behavior of athermal polymer/nanoparticle blends near a hard substrate. We apply the density functional theory of Tripathi and Chapman to these blends. We find a first order phase transition where the nanoparticles expel the polymer from the surface to form a monolayer. The transition density depends on the polymer length and the system bulk density. The effect is due to the packing entropy of the species and configurational entropy of the polymer. The simplicity of the system allows us to understand the so-called ``entropic-push'' observed in experiments.
△ Less
Submitted 31 October, 2007;
originally announced October 2007.
-
Stability of cylindrical domains in phase-separating binary fluids in shear flow
Authors:
Amalie Frischknecht
Abstract:
The stability of a long cylindrical domain in a phase-separating binary fluid in an external shear flow is investigated by linear stability analysis. Using the coupled Cahn-Hilliard and Stokes equations, the stability eigenvalues are derived analytically for long wavelength perturbations, for arbitrary viscosity contrast between the two phases. The shear flow is found to suppress and sometimes c…
▽ More
The stability of a long cylindrical domain in a phase-separating binary fluid in an external shear flow is investigated by linear stability analysis. Using the coupled Cahn-Hilliard and Stokes equations, the stability eigenvalues are derived analytically for long wavelength perturbations, for arbitrary viscosity contrast between the two phases. The shear flow is found to suppress and sometimes completely stabilize both the hydrodynamic Rayleigh instability and the thermodynamic instability of the cylinder against varicose perturbations, by mixing with nonaxisymmetric perturbations. The results are consistent with recent observations of a ``string phase'' in phase-separating fluids in shear.
△ Less
Submitted 3 June, 1998; v1 submitted 15 May, 1998;
originally announced May 1998.
-
Effect of Shear Flow on the Stability of Domains in Two Dimensional Phase-Separating Binary Fluids
Authors:
Amalie Frischknecht
Abstract:
We perform a linear stability analysis of extended domains in phase-separating fluids of equal viscosity, in two dimensions. Using the coupled Cahn-Hilliard and Stokes equations, we derive analytically the stability eigenvalues for long wavelength fluctuations. In the quiescent state we find an unstable varicose mode which corresponds to an instability towards coarsening. This mode is stabilized…
▽ More
We perform a linear stability analysis of extended domains in phase-separating fluids of equal viscosity, in two dimensions. Using the coupled Cahn-Hilliard and Stokes equations, we derive analytically the stability eigenvalues for long wavelength fluctuations. In the quiescent state we find an unstable varicose mode which corresponds to an instability towards coarsening. This mode is stabilized when an external shear flow is imposed on the fluid. The effect of the shear is seen to be qualitatively similar to that found in experiments.
△ Less
Submitted 18 July, 1997;
originally announced July 1997.
