Self-attention approach for crop disease prediction
13 Novembre 2021
Catégorie : Stagiaire
Title “Self-attention approach for crop disease prediction”
Crop diseases cause economic losses, reduced crop quality and yields, negative environmental impact when using crop protection products to treat them. The detection and management of crop diseases is therefore a major challenge for agriculture as well as for the economy and the environment. New modern technologies, such as IoT, drones, remote sensing, big data, and artificial intelligence, integrated by information and communication technologies (ICT), have triggered a new era for digital agronomy. Indeed, these technologies offer enormous potential to solve challenging problems such as early disease detection and management. With the state of the art showing a trend towards their widespread application. On the other hand, advances in sensor technology have raised many challenges, such as processing and analysis of heterogeneous and noisy data, security, reliability, etc.
The work proposed here consists in developing and evaluating algorithms based on the principles of convolutional networks, recurrent networks, transformers... with self-attention approach applied to vineyard agronomic data. In a first step, the work will focus on the construction of models for disease assessment in 2D multispectral images. In a second step, we will use other type of data for disease prediction.
Student in Master 2 or final year of engineering school in the field of computer science. Good writing skill and a good level in English language. Experience in machine learning and development will be appreciated.
CV, transcript and letter of motivation to be sent to : firstname.lastname@example.org , email@example.com
INSA CVL, Laboratoire PRSIME, campus de Bourges