Intermittent Inference with Nonuniformly Compressed Multi-Exit Neural Network for Energy Harvesting Powered Devices

Abstract

This work aims to enable persistent, event-driven sensing and decision capabilities for energy-harvesting (EH)-powered devices by deploying lightweight DNNs onto EH-powered devices. However, harvested energy is usually weak and unpredictable and even lightweight DNNs take multiple power cycles to finish one inference. To eliminate the indefinite long wait to accumulate energy for one inference and to optimize the accuracy, we developed a power trace-aware and exit-guided network compression algorithm to compress and deploy multi-exit neural networks to EH-powered microcontrollers (MCUs) and select exits during execution according to available energy. The experimental results show superior accuracy and latency compared with state-of-the-art techniques.

References in corpus (6)

• Continuous control with deep reinforcement learning
• AMC: AutoML for Model Compression and Acceleration on Mobile Devices
• Pruning Filters for Efficient ConvNets
• SpArSe: Sparse Architecture Search for CNNs on Resource-Constrained Microcontrollers
• On Neural Architecture Search for Resource-Constrained Hardware Platforms
• NASS: Optimizing Secure Inference via Neural Architecture Search

Cited by in corpus (3)

• Standing on the Shoulders of Giants: Hardware and Neural Architecture Co-Search with Hot Start
• Enabling On-Device CNN Training by Self-Supervised Instance Filtering and Error Map Pruning
• Enabling On-Device Self-Supervised Contrastive Learning With Selective Data Contrast