OscNet v1.5: Energy Efficient Hopfield Network on CMOS Oscillators for Image Classification
This addresses energy efficiency for edge computing applications, though it is incremental as it builds on existing oscillator hardware and Hopfield networks.
The paper tackles the problem of high computational energy costs in machine learning by proposing a Hopfield Network algorithm implemented on CMOS Oscillator Networks (OscNet), achieving an 8% accuracy improvement on MNIST and using only 24% of connections with a 0.1% accuracy drop.
Machine learning has achieved remarkable advancements but at the cost of significant computational resources. This has created an urgent need for a novel and energy-efficient computational fabric and corresponding algorithms. CMOS Oscillator Networks (OscNet) is a brain inspired and specially designed hardware for low energy consumption. In this paper, we propose a Hopfield Network based machine learning algorithm that can be implemented on OscNet. The network is trained using forward propagation alone to learn sparsely connected weights, yet achieves an 8% improvement in accuracy compared to conventional deep learning models on MNIST dataset. OscNet v1.5 achieves competitive accuracy on MNIST and is well-suited for implementation using CMOS-compatible ring oscillator arrays with SHIL. In oscillator-based inference, we utilize only 24% of the connections used in a fully connected Hopfield network, with merely a 0.1% drop in accuracy. OscNet v1.5 relies solely on forward propagation and employs sparse connections, making it an energy-efficient machine learning pipeline designed for oscillator computing fabric. The repository for OscNet family is: https://github.com/RussRobin/OscNet .