SOI: Scaling Down Computational Complexity by Estimating Partial States of the Model

TL;DR

This paper introduces SOI, a novel method that reduces the computational complexity of neural networks in embedded systems by estimating partial states, leveraging data continuity to enable faster inference without full model recalculations.

Contribution

The paper presents SOI, a new approach that estimates partial states of neural networks to significantly decrease computational load in resource-constrained devices.

Findings

SOI achieves faster inference times in embedded neural networks.

SOI maintains accuracy while reducing computation.

The method exploits data continuity for efficient extrapolation.

Abstract

Consumer electronics used to follow the miniaturization trend described by Moore's Law. Despite increased processing power in Microcontroller Units (MCUs), MCUs used in the smallest appliances are still not capable of running even moderately big, state-of-the-art artificial neural networks (ANNs) especially in time-sensitive scenarios. In this work, we present a novel method called Scattered Online Inference (SOI) that aims to reduce the computational complexity of ANNs. SOI leverages the continuity and seasonality of time-series data and model predictions, enabling extrapolation for processing speed improvements, particularly in deeper layers. By applying compression, SOI generates more general inner partial states of ANN, allowing skipping full model recalculation at each inference.