Stage Ingénieur/Master 2: Hyperspectral Image description using deep learning
XLIM/University of Poitiers
Contact : email@example.com
Traditional RGB imaging have been widely used for remote sensing scene retrieval and classification . However, such image representation lacks precision that profiles the physical proprieties of a scene. Hyperspectral imaging (HSI) is an emerging imaging modality that contains hundreds of spectral bands (varying from the visible to the infrared ranges) allowing profile materials and organisms that only hyperspectral sensors can provide (Example: AVIRIS sensors (https://aviris.jpl.nasa.gov/)). With the emergence of huge volumes of high-resolution hyperspectral images produced by different types of imaging sensors, analyzing and retrieving of these images requires effective image description and quantification techniques.In this internship, we will investigate the development of deep learning approaches, already used for RGB remote sensing data , for space-spectral Hyperspectral images description.The challenge is to quantify the spectral and spatial content of an HSI image in a compact signature for efficient image retrieval.
The objective of this internship is to study and implement new solutions for HSI image description and quantification for content-based HSI retrieval using recent development in deep learning approaches  then compare the obtained performance with an RGB-based retrieval system. The proposed method will be tested on the HSI-XLIM dataset .
Keywords: Hyperspectral Imagining, deep learning, content-based retrieval
Location: XLIM laboratory, University of Poitiers, France
Duration: 6 months.
Starting date: February/March 2019.
Salary: ~ 500 euros/month. In addition to the salary, we propose to pay the accommodation cost (in a student residence of the CROUS of Poitiers) for the student who is not coming from the University of Poitiers.
WeixunZhouaShawnNewsambCongminLia, ZhenfengShaoa, PatternNet: A benchmark dataset for performance evaluation of remote sensing image retrieval, ISPRS Journal of Photogrammetry and Remote Sensing, Volume 145, Part A, November 2018, Pages 197-209
 Gui-Song Xia, Xin-Yi Tong, +3 authors Liangpei Zhang Exploiting Deep Features for Remote Sensing Image Retrieval: A Systematic Investigation, Published 2017 in ArXiv (https://arxiv.org/abs/1707.07321)
 Olga Russakovsky*, Jia Deng*, Hao Su, Jonathan Krause, Sanjeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael Bernstein, Alexander C. Berg and Li Fei-Fei.ImageNet Large Scale Visual Recognition Challenge. IJCV, 2015.
 Olfa Ben-Ahmed, Thierry Urruty, Noel Richard, Christine Fernandez-Maloigne, “ A Study on the Sensitivity-Based Discriminative Hyperspectral Image Content Representation” International Conference on Content-Based Multimedia Indexing (CBMI) 2018
Strong knowledge in image processing, machine learning (deep learning) are required. Knowledge of at least one platform for neural network processing and related programming tool python is required (Keras/TensorFlow …)
(c) GdR 720 ISIS - CNRS - 2011-2018.