Description

This course provides an overview of advanced techniques for computer vision, either 2D or 3D, either static or dynamic. Methods mostly aim at extracting relevant information from the observed scene. The course includes lessons and practical work.

Prerequisites for Participation

• Specific prerequisites: Good knowledge of the basics of image processing, geometry
• Recommended prerequisites: Good level in applied mathematics and computer science.
• Programming language: Python, C, C++, Matlab

Intended Learning Outcomes

At the end of this course, the students will have advanced knowledge and skills in various theories of image processing and analysis. They will be able to solve theoretical problems, as well as applied ones.

Content

• Biological perception of 3D depth and motion
• Co-design for 3D vision and motion analysis
• Projective geometry
• Fundamentals matrices, RANSAC
• Disparity maps
• Maximal flow in graphs
• Stereo-vision
• Motion estimation
• Motion detection, tracking, action modeling
• Faces detection

Assessment and Grading Procedures

• Written examen and evaluation of pratical works
• Examiners: Antoine Manzanera, Pascal Monasse, Kevin Bailly, Dominique Béréziat

Workload calculation (contact hours, homework, exam preparation,..)

• 4h weekly contact hours x 15 weeks = 60 h
• 4h weekly hours preparation and afterwork x 15 weeks = 60 h
• Exams preparation: 30 h

Recommended Reading, Course Material

• Computer Vision: Algorithms and Applications, by Richard Szeliski, Springer Science & Business Media, 2010
• Three-Dimensional Computer Vision, by Olivier Faugeras, MIT press, 1993.