Furcifer: a Context Adaptive Middleware for Real-world Object Detection Exploiting Local, Edge, and Split Computing in the Cloud Continuum

• Published in: Proceedings of the IEEE International Conference on Pervasive Computing and Communications pp. 47 - 56
• Main Authors: Mendula, Matteo, Bellavista, Paolo, Levorato, Marco, de Guevara Contreras, Sharon Ladron
• Format: Conference Proceeding
• Language: English
• Published: IEEE 11.03.2024
• Subjects: Artificial neural networks; Computational modeling; Edge Computing; Energy consumption; Image compression; Image edge detection; Machine Learning; Mobile agents; Object detection; Real-time systems; Wireless communication
• ISSN: 2474-249X
• DOI: 10.1109/PerCom59722.2024.10494426

Modern real-time applications widely embed compute intense neural algorithms at their core. Current solutions to support such algorithms either deploy highly-optimized Deep Neural Networks at mobile devices or offload the execution of possibly larger higher-performance neural models to edge servers. While the former solution typically maps to higher energy consumption and lower performance, the latter necessitates the low-latency wireless transfer of high volumes of data. Time-varying variables describing the state of these systems, such as connection quality and system load, determine the optimality of the different computing configurations in terms of energy consumption, task performance, and latency. Herein, we propose Furcifer, a framework capable of dynamically adapting the cloud continuum computing configuration in response to the perceived state of the system. Our container-based approach incorporates low-complexity predictors that generalize well across operating environments. In addition, we develop a highly optimized split Deep Neural Network model, which achieves in-model supervised compression and enhances task offloading. Experimental results for object detection across diverse conditions, environments, and wireless technologies, show Furcifer's remarkable outcomes, including a 2x energy reduction, 30% higher mean Average Precision score than pure local computing, and a notable three-fold increase in frame per second rate compared to static offloading.