Métois, Mr Gabriel (2023) Estimation de la traversabilité d’un terrain par apprentissage profond PRE - Research Project, ENSTA.
Full text not available from this repository.
Abstract
In the context of robotic navigation in outdoor, unstructured terrain, it is necessary to be able to estimate the difficulty of traversing a terrain when planning a trajectory. This machine learning approach uses a two-network self-supervised learning scheme. While the first is used to form the training set for the second, the aim here is to improve the performance of the second and develop the associated measurement and software integration tools.
Item Type: | Thesis (PRE - Research Project) |
---|---|
Uncontrolled Keywords: | Robotique extérieure, ROS, Réseau de Neurone, Analyse de terrain |
Subjects: | Information and Communication Sciences and Technologies |
ID Code: | 9587 |
Deposited By: | Gabriel MÉTOIS |
Deposited On: | 24 août 2023 15:02 |
Dernière modification: | 24 août 2023 15:02 |
Repository Staff Only: item control page