1. MULTI-BRANCH DEEP LEARNING MODEL FOR DETECTION OF SETTLEMENTS WITHOUT ELECTRICITY
- Author
-
Thomas Di Martino, Elise Colin Koeniguer, Maxime Lenormand, DTIS, ONERA, Université Paris Saclay [Palaiseau], ONERA-Université Paris-Saclay, CentraleSupélec, and GREC, christine
- Subjects
010504 meteorology & atmospheric sciences ,Computer science ,[SPI] Engineering Sciences [physics] ,Computation ,Feature extraction ,0211 other engineering and technologies ,Multi Sensor ,02 engineering and technology ,[MATH] Mathematics [math] ,[INFO] Computer Science [cs] ,computer.software_genre ,01 natural sciences ,[PHYS] Physics [physics] ,Remote Sensing ,[SPI]Engineering Sciences [physics] ,Deep Learning ,[INFO]Computer Science [cs] ,[MATH]Mathematics [math] ,Representation (mathematics) ,021101 geological & geomatics engineering ,0105 earth and related environmental sciences ,[PHYS]Physics [physics] ,business.industry ,Deep learning ,Sensor fusion ,Classification ,Data Fusion Contest ,Task (computing) ,Multi Temporal ,Feature (computer vision) ,Data mining ,Artificial intelligence ,F1 score ,business ,computer - Abstract
International audience; We introduce a multi-branch Deep Learning architecture that allows for the extraction of multi-scale features. Exploiting the data multi-modality structure through the combined use of various feature extractors provides high performance on data fusion tasks. Furthermore, the representation of the multitemporality of the data using sensor-specific 3D convolutions with custom kernel size extracts temporal features at an early computation stage. Our methodology allows reaching performance up to 0.8876 F1 Score on the development phase dataset and around 0.8798 on the test phase dataset. Finally, we demonstrate the contribution of each sensor to the prediction task with the design of data-focused experiments.
- Published
- 2021