TOWARDS DISTILLATION OF DEEP NEURAL NETWORKS FOR SATELLITE ON-BOARD IMAGE SEGMENTATION

Cubesats platforms expansion increases the need to simplify payloads and to optimize downlink data capabilities. A promising solution is to enhance on-board software, in order to take early decisions, automatically. However, the most efficient methods for data analysis are generally large deep neura...

Full description

Saved in:
Bibliographic Details
Main Authors: de Vieilleville, F., Lagrange, A., Ruiloba, R., May, S.
Format: Conference Proceeding
Language:English
Subjects:
Accuracy
Artificial neural networks
Cubesat
Data analysis
Decision analysis
Distillation
Distilling
Image segmentation
Inference
Neural networks
Payloads
Satellite imagery
Ships
Citations: Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Cubesats platforms expansion increases the need to simplify payloads and to optimize downlink data capabilities. A promising solution is to enhance on-board software, in order to take early decisions, automatically. However, the most efficient methods for data analysis are generally large deep neural networks (DNN) oversized to be loaded and processed on limited hardware capacities of cubesats. To use them, we must reduce the size of DNN while accommodating efficiency in terms of both accuracy and inference cost. In this paper, we propose a distillation method which reduces image segmentation deep neural network’s size to fit into on board processors. This method is presented through a ship detection example comparing accuracy and inference costs for several networks.
ISSN:2194-9034
1682-1750
2194-9034
DOI:10.5194/isprs-archives-XLIII-B2-2020-1553-2020