Infer-EDGE: Dynamic DNN Inference Optimization in 'Just-in-time' Edge-AI Implementations

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Mounesan, Motahare, Zhang, Xiaojie, Debroy, Saptarshi
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909470989221888
author Mounesan, Motahare
Zhang, Xiaojie
Debroy, Saptarshi
author_facet Mounesan, Motahare
Zhang, Xiaojie
Debroy, Saptarshi
contents Balancing mutually diverging performance metrics, such as end-to-end latency, accuracy, and device energy consumption, is a challenging undertaking for deep neural network (DNN) inference in Just-in-Time edge environments that are inherently resource-constrained and loosely coupled. In this paper, we design and develop the Infer-EDGE framework that seeks to strike such a balance for latency-sensitive video processing applications. First, using comprehensive benchmarking experiments, we develop intuitions about the trade-off characteristics, which are then used by the framework to develop an Advantage Actor-Critic (A2C) Reinforcement Learning (RL) approach that can choose optimal run-time DNN inference parameters aligning the performance metrics based on the application requirements. Using real-world DNNs and a hardware testbed, we evaluate the benefits of the Infer-EDGE framework in terms of device energy savings, inference accuracy improvement, and end-to-end inference latency reduction.
format Preprint
id arxiv_https___arxiv_org_abs_2501_18842
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Infer-EDGE: Dynamic DNN Inference Optimization in 'Just-in-time' Edge-AI Implementations
Mounesan, Motahare
Zhang, Xiaojie
Debroy, Saptarshi
Distributed, Parallel, and Cluster Computing
Balancing mutually diverging performance metrics, such as end-to-end latency, accuracy, and device energy consumption, is a challenging undertaking for deep neural network (DNN) inference in Just-in-Time edge environments that are inherently resource-constrained and loosely coupled. In this paper, we design and develop the Infer-EDGE framework that seeks to strike such a balance for latency-sensitive video processing applications. First, using comprehensive benchmarking experiments, we develop intuitions about the trade-off characteristics, which are then used by the framework to develop an Advantage Actor-Critic (A2C) Reinforcement Learning (RL) approach that can choose optimal run-time DNN inference parameters aligning the performance metrics based on the application requirements. Using real-world DNNs and a hardware testbed, we evaluate the benefits of the Infer-EDGE framework in terms of device energy savings, inference accuracy improvement, and end-to-end inference latency reduction.
title Infer-EDGE: Dynamic DNN Inference Optimization in 'Just-in-time' Edge-AI Implementations
topic Distributed, Parallel, and Cluster Computing
url https://arxiv.org/abs/2501.18842