Data analysis and applications in signal and image processing


Head: Saïd MOUSSAOUI

Objectives

The objective of this specialisation is to train multidisciplinary engineers to design and implement methodological and algorithmic solutions to data processing problems in various industrial application fields.



Courses are based on the theory and the practice of methods from computational statistics, applied mathematics, signal and image processing, as well as applied computer science and scientific computing. These courses also offer application-oriented content from healthcare, research and development, imaging science, information and communication technology.



This specialisation confers Centrale engineers the skills needed for a professional orientation to research and innovation in industrial and academic fields related to data sciences, audio engineering, industrial imaging, computer-aided decision and biomedical engineering.

List of Courses

 

Autumn Semester

 
  • Scientific computing and numerical optimization
  • Signal representation and analysis
  • Image processing and analysis
  • Statistical data modelling and analysis
  • Machine learning
  • System identification and signal filtering
  • Computational imaging and inverse methods
  • Biomedical signal analysis
  • Project in signal and image processing
 

Spring Semester
 

  • Biomedical imaging
  • Audio content analysis and Information Retrieval
  • Multi-sensor data analysis
  • R&D applications
  • Project in signal and image processing

 


 

Examples of past projects

  • Joint detection-estimation of hymodynamic responses from functional MRI data
  • Ultrasound data analysis for depth estimation in non-destructive testing
  • Convex k-means clustering from approximate pairwise comparisons
  • Unsupervised data clustering for acoustic quality assessment in urban areas
  • Implementation of deep learning algorithms for CT scan image segmentation
  • Optimization of a Brain-Computer Interface including a Virtual Reality feedback
  • Online acquisition and unmixing of hyperspectral images

Examples of past internships

  • Short-term audio source separation filter estimation from recurrent convolutional networks (Orange, Rennes)
  • Automatic detection of vine rows on aerial images (Avion Jaune, Paris)
  • Predicting the risk of delays for the operation of major train stations (SNCF, Paris)
  • Correlation between foot and hand movement in pedestrian navigation (IFSTTAR, Nantes)
  • Machine learning on biomedical images (CHU Nantes)
  • Evaluation of the mental load induced by a brain-computer interface system coupled with virtual reality (OnePoint and CHU Nantes)
  • Development of an image processing algorithm for the correction of artifacts in 2D and 3D mammography acquisitions (GE Healthcare, The Netherlands)

Sectors of activity and employment prospects

The expertise acquired through this specialisation affords opportunities for engineers across a broad range of industrial sectors: ICT, biomedical engineering, sound and multi-media as well as the automotive and aeronautics sectors.
Published on November 2, 2015 Updated on July 7, 2020