KIT | KIT-Bibliothek | Impressum | Datenschutz

Hardware-aware Partitioning of Convolutional Neural Network Inference for Embedded AI Applications

Kreß, Fabian ORCID iD icon 1; Hoefer, Julian ORCID iD icon 1; Hotfilter, Tim ORCID iD icon 1; Walter, Iris ORCID iD icon 1; Sidorenko, Vladimir 1; Harbaum, Tanja 1; Becker, Jürgen 1
1 Institut für Technik der Informationsverarbeitung (ITIV), Karlsruher Institut für Technologie (KIT)

Abstract (englisch):

Embedded image processing applications like multicamera-based object detection or semantic segmentation are often based on Convolutional Neural Networks (CNNs) to provide precise and reliable results. The deployment of CNNs in embedded systems, however, imposes additional constraints such as latency restrictions and limited energy consumption in the sensor platform. These requirements have to be considered during hardware/software co-design of embedded Artifical Intelligence (AI) applications. In addition, the transmission of uncompressed image data from the sensor to a central edge node requires large bandwidth on the link, which must also be taken into account during the design phase.Therefore, we present a simulation toolchain for fast evaluation of hardware-aware CNN partitioning for embedded AI applications. This approach explores an efficient workload distribution between sensor nodes and a central edge node. Neither processing all layers close to the sensor nor transmitting all uncompressed raw data to the edge node is an optimal solution for each use case. Hence, our proposed simulation toolchain evaluates power and performance metrics for each reasonable partitioning point in a CNN. ... mehr

Postprint §
DOI: 10.5445/IR/1000150617
Veröffentlicht am 13.09.2023
DOI: 10.1109/DCOSS54816.2022.00034
Zitationen: 3
Zitationen: 2
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Technik der Informationsverarbeitung (ITIV)
Publikationstyp Proceedingsbeitrag
Publikationsdatum 12.09.2022
Sprache Englisch
Identifikator KITopen-ID: 1000150617
Erschienen in 18th International Conference on Distributed Computing in Sensor Systems (DCOSS)
Veranstaltung 18th International Conference on Distributed Computing in Sensor Systems (DCOSS 2022), Los Angeles, CA, USA, 30.05.2022 – 01.06.2022
Verlag IEEEXplore
Seiten 133-140
Nachgewiesen in Scopus
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page