Energy-Efficient Adaptive Machine Learning on IoT End-Nodes With Class-Dependent Confidence
This work addresses energy efficiency for IoT applications, but it is incremental as it builds on existing sequential model methods.
The paper tackled the problem of energy-efficient machine learning on IoT devices by proposing a per-class threshold method for early stopping in sequential models, showing it significantly reduces energy consumption compared to a single-threshold approach.
Energy-efficient machine learning models that can run directly on edge devices are of great interest in IoT applications, as they can reduce network pressure and response latency, and improve privacy. An effective way to obtain energy-efficiency with small accuracy drops is to sequentially execute a set of increasingly complex models, early-stopping the procedure for "easy" inputs that can be confidently classified by the smallest models. As a stopping criterion, current methods employ a single threshold on the output probabilities produced by each model. In this work, we show that such a criterion is sub-optimal for datasets that include classes of different complexity, and we demonstrate a more general approach based on per-classes thresholds. With experiments on a low-power end-node, we show that our method can significantly reduce the energy consumption compared to the single-threshold approach.