核心概念
REDS introduces structured sparsity to adapt deep models to variable resources efficiently.
統計
State-of-the-art machine learning pipelines generate resource-agnostic models, not capable to adapt at runtime.
In response to dynamic resource constraints, REDS demonstrate an adaptation time of under 40µs utilizing a 2-layer fully-connected network on Arduino Nano 33 BLE.
引用
"Deep models deployed on edge devices frequently encounter resource variability."
"In contrast to the state-of-the-art, REDS use structured sparsity constructively by exploiting permutation invariance of neurons."