Artificial Vision
A.Y. 2025/2026
Learning objectives
The aim of the course is to provide students with the skills to analyze images, videos, and three-dimensional reconstructions, in order to extract semantic meaning by interpreting the real world through the recognition of objects, subjects, actions, and situations depicted in the scene. The skills will focus on the fundamentals of vision and Deep Learning techniques.
Expected learning outcomes
- Understand the fundamental principles of image formation
- Learn techniques for reconstructing three-dimensional models of real-world objects
- Gain knowledge of learning methods—particularly Deep Learning techniques—used in computer vision to identify and recognize objects or actions from images and videos
- Learn techniques for reconstructing three-dimensional models of real-world objects
- Gain knowledge of learning methods—particularly Deep Learning techniques—used in computer vision to identify and recognize objects or actions from images and videos
Lesson period: Third four month period
Assessment methods: Esame
Assessment result: voto verbalizzato in trentesimi
Single course
This course can be attended as a single course.
Course syllabus and organization
Single session
Responsible
Lesson period
Third four month period
Course syllabus
Aim of this course is to examine the fundamental concepts in the field of computer vision, organized in the following modules:
1) Image formation
- Geometric camera models
- Camera Calibration
- Brightness models
- Color models
2) "Early" vision
- Linear Filtering
- Feature extraction
- Stereopsis (binocular vision)
- Structure from Motion (multi-view)
- 3D Registration
3) Machine learning
- Deep Learning Computation
- Convolutional Neural Networks
- Transformers
- Recurrent Neural networks
- Graph Neural Networks
- Generative models
- Self-Supervised Learning
1) Image formation
- Geometric camera models
- Camera Calibration
- Brightness models
- Color models
2) "Early" vision
- Linear Filtering
- Feature extraction
- Stereopsis (binocular vision)
- Structure from Motion (multi-view)
- 3D Registration
3) Machine learning
- Deep Learning Computation
- Convolutional Neural Networks
- Transformers
- Recurrent Neural networks
- Graph Neural Networks
- Generative models
- Self-Supervised Learning
Prerequisites for admission
It is recommended to be familiar with the basics of:
- probability and statistics
- image processing
- signal processing
- probability and statistics
- image processing
- signal processing
Teaching methods
Lectures and demo in Python
Teaching Resources
Web site:
https://github.com/lanzarotti/
Class material:
- Lecture slides
- Software code for exercises
Books:
- [Modules 1 and 2] D.A. Forsyth, J. Ponce - Computer Vision - A Modern Approach - Pearson, 2nd edition
- [Module 3] C.M. Bishop, H. Bishop - Deep Learning - Foundations and Concepts - Springer
- Optional - Programming resource: A. Zhang, Z.C. Lipton, M. Li, A.J. Smola - Dive into Deep Learning
https://github.com/lanzarotti/
Class material:
- Lecture slides
- Software code for exercises
Books:
- [Modules 1 and 2] D.A. Forsyth, J. Ponce - Computer Vision - A Modern Approach - Pearson, 2nd edition
- [Module 3] C.M. Bishop, H. Bishop - Deep Learning - Foundations and Concepts - Springer
- Optional - Programming resource: A. Zhang, Z.C. Lipton, M. Li, A.J. Smola - Dive into Deep Learning
Assessment methods and Criteria
The exam consists of a written test with open-ended questions or a project, which involves submitting a report and the produced code, as well as presenting the project itself.
The evaluation is expressed on a thirty-point scale.
The evaluation is expressed on a thirty-point scale.
Professor(s)