Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

E-QUARTIC: Energy Efficient Edge Ensemble of Convolutional Neural Networks for Resource-Optimized Learning

Le Zhang, Onat Gungor, Flavio Ponzina, Tajana Rosing

TL;DR

This work proposes E-QUARTIC, a novel Energy Efficient Edge Ensembling framework to build ensembles of CNNs targeting Artificial Intelligence (AI)-based embedded systems, and shows that the proposed design enables concurrent on-device training and high-quality inference execution at the edge, limiting the performance and energy overheads to less than 0.04%.

Abstract

Ensemble learning is a meta-learning approach that combines the predictions of multiple learners, demonstrating improved accuracy and robustness. Nevertheless, ensembling models like Convolutional Neural Networks (CNNs) result in high memory and computing overhead, preventing their deployment in embedded systems. These devices are usually equipped with small batteries that provide power supply and might include energy-harvesting modules that extract energy from the environment. In this work, we propose E-QUARTIC, a novel Energy Efficient Edge Ensembling framework to build ensembles of CNNs targeting Artificial Intelligence (AI)-based embedded systems. Our design outperforms single-instance CNN baselines and state-of-the-art edge AI solutions, improving accuracy and adapting to varying energy conditions while maintaining similar memory requirements. Then, we leverage the multi-CNN structure of the designed ensemble to implement an energy-aware model selection policy in energy-harvesting AI systems. We show that our solution outperforms the state-of-the-art by reducing system failure rate by up to 40% while ensuring higher average output qualities. Ultimately, we show that the proposed design enables concurrent on-device training and high-quality inference execution at the edge, limiting the performance and energy overheads to less than 0.04%.

E-QUARTIC: Energy Efficient Edge Ensemble of Convolutional Neural Networks for Resource-Optimized Learning

TL;DR

This work proposes E-QUARTIC, a novel Energy Efficient Edge Ensembling framework to build ensembles of CNNs targeting Artificial Intelligence (AI)-based embedded systems, and shows that the proposed design enables concurrent on-device training and high-quality inference execution at the edge, limiting the performance and energy overheads to less than 0.04%.

Abstract

Ensemble learning is a meta-learning approach that combines the predictions of multiple learners, demonstrating improved accuracy and robustness. Nevertheless, ensembling models like Convolutional Neural Networks (CNNs) result in high memory and computing overhead, preventing their deployment in embedded systems. These devices are usually equipped with small batteries that provide power supply and might include energy-harvesting modules that extract energy from the environment. In this work, we propose E-QUARTIC, a novel Energy Efficient Edge Ensembling framework to build ensembles of CNNs targeting Artificial Intelligence (AI)-based embedded systems. Our design outperforms single-instance CNN baselines and state-of-the-art edge AI solutions, improving accuracy and adapting to varying energy conditions while maintaining similar memory requirements. Then, we leverage the multi-CNN structure of the designed ensemble to implement an energy-aware model selection policy in energy-harvesting AI systems. We show that our solution outperforms the state-of-the-art by reducing system failure rate by up to 40% while ensuring higher average output qualities. Ultimately, we show that the proposed design enables concurrent on-device training and high-quality inference execution at the edge, limiting the performance and energy overheads to less than 0.04%.
Paper Structure (8 sections, 2 equations, 5 figures, 3 tables)

This paper contains 8 sections, 2 equations, 5 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. E-QUARTIC Generation
  4. E-QUARTIC Adaptation to Dynamic Energy Environments
  5. Evaluation
  6. Experimental Setup
  7. Results
  8. Conclusion

Figures (5)

  • Figure 1: E-QUARTIC Overview
  • Figure 2: Ratio of multiply-accumulate operations (MACs) for each layer of ResNet-8 (Baseline) and an E-QUARTIC design including two CNN instances.
  • Figure 3: Concurrent training and inference stages in low and high energy conditions, for $N=4$ and $k=3$.
  • Figure 4: Mean accuracy-failure rate reduction Pareto graphs
  • Figure 5: Harvested power trace (top), battery levels and inference traces of ResNet-8 banbury2021mlperf (middle) and the corresponding E-QUARTIC implementation (bottom).