Low-Precision Floating-Point for Efficient On-Board Deep Neural Network Processing - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Low-Precision Floating-Point for Efficient On-Board Deep Neural Network Processing

Résumé

One of the major bottlenecks in high-resolution Earth Observation (EO) space systems is the downlink between the satellite and the ground. Due to hardware limitations, onboard power limitations or ground-station operation costs, there is a strong need to reduce the amount of data transmitted. Various processing methods can be used to compress the data. One of them is the use of on-board deep learning to extract relevant information in the data. However, most ground-based deep neural network parameters and computations are performed using single-precision floating-point arithmetic, which is not adapted to the context of on-board processing. We propose to rely on quantized neural networks and study how to combine low precision (mini) floating-point arithmetic with a Quantization-Aware Training methodology. We evaluate our approach with a semantic segmentation task for ship detection using satellite images from the Airbus Ship dataset. Our results show that 6-bit floating-point quantization for both weights and activations can compete with single-precision without significant accuracy degradation. Using a Thin U-Net 32 model, only a 0.3% accuracy degradation is observed with 6-bit minifloat quantization (a 6-bit equivalent integer-based approach leads to a 0.5% degradation). An initial hardware study also confirms the potential impact of such lowprecision floating-point designs, but further investigation at the scale of a full inference accelerator is needed before concluding whether they are relevant in a practical on-board scenario.
Fichier principal
Vignette du fichier
Low-Precision_Floating-Point_for_Efficient_On-Board_Deep_Neural_Network_Processing.pdf (1.14 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04252197 , version 1 (20-10-2023)

Licence

Paternité

Identifiants

Citer

Cédric Gernigon, Silviu-Ioan Filip, Olivier Sentieys, Clément Coggiola, Mickaël Bruno. Low-Precision Floating-Point for Efficient On-Board Deep Neural Network Processing. European Data Handling & Data Processing Conference (EDHPC), European Space Agency (ESA), Oct 2023, Juan-Les-Pins, France. pp.1-8, ⟨10.23919/EDHPC59100.2023.10396014⟩. ⟨hal-04252197⟩
58 Consultations
102 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More