Pubblicazioni

Autori:

Capogrosso, Luigi; Fraccaroli, Enrico; Cristani, Marco; Fummi, Franco; Chakraborty, Samarjit

Titolo:

LO-SC: Local-Only Split Computing for Accurate Deep Learning on Edge Devices

Anno:

2025

Tipologia prodotto:

Contributo in atti di convegno

Tipologia ANVUR:

Contributo in Atti di convegno

Lingua:

Inglese

Formato:

Elettronico

Titolo del Convegno:

38th International Conference on VLSI Design (VLSID)

Luogo:

Bangalore, India

Periodo:

04-08 January 2025

Intervallo pagine:

445-450

Parole chiave:

Deep Neural Networks; Edge Device

Breve descrizione dei contenuti:

Split Computing (SC) enables deploying a Deep Neural Network (DNN) on edge devices with limited resources by splitting the workload between the edge device and a remote server. However, relying on a server can be expensive, requires a reliable network, and introduces unpredictable latency. Existing solutions for on-device DNNs deployment often sacrifice accuracy for efficiency. In this paper, we study how to use the concepts from SC to split a DNN for executing on the same device without compromising accuracy. In other words, we propose Local-Only Split Computing (LO-SC), a new approach to split a DNN for execution entirely on the edge device while maintaining high accuracy and predictable latency. We formalize LO-SC as a MixedInteger Linear Problem (MILP) problem and solve it using a multi-constrained ordered knapsack algorithm. The proposed method achieves promising results on both synthetic and realworld data, offering a viable alternative for accurately deploying DNNs on resource-constrained edge devices. The source code is available at https://github.com/intelligolabs/LO-SC.

Pagina Web:

https://ieeexplore.ieee.org/abstract/document/10900702

Id prodotto:

144572

Handle IRIS:

11562/1156847

ultima modifica:

11 marzo 2025

Citazione bibliografica:

Capogrosso, Luigi; Fraccaroli, Enrico; Cristani, Marco; Fummi, Franco; Chakraborty, Samarjit, LO-SC: Local-Only Split Computing for Accurate Deep Learning on Edge Devices  in Proceedings of the 38th International Conference on VLSI Design (VLSID) , Atti di "38th International Conference on VLSI Design (VLSID)" , Bangalore, India , 04-08 January 2025 , 2025 , pp. 445-450

Consulta la scheda completa presente nel repository istituzionale della Ricerca di Ateneo