Course Note: Tübingen-CVML-Introduction
Introduction of AI and Computer Vision
Credits: Tübingen Machine Learning | Computer Vision - Andreas Geiger
Computer Vision - Lecture 1.2 (Introduction: Introduction) - YouTube
Computer Vision - Lecture 1.3 (Introduction: History of Computer Vision) - YouTube
Overview of AI
An attempt will be made to find how to make machines use language, form abstractions and concepts, solve kinds of problems now reserved for humans, and improve themselves.
[John McCarthy]
Artificial Intelligence
Artificial Intelligence contained multiple fields:
- Machine Learning
- Computer Vision
- Computer Graphics
- Natural Language Processing
- Robotics & Control...
Computer Vision
Computer Vision
Goal of Computer Vision is to convert light into meaning (geometric, semantic,...)
- What does it mean, to visual perception?
- To discover from images what is present in the world, where things, what actions are taking place, to predict and anticipate events in the world.
Computer Vision Applications
- Robotics
- Medical applications
- 3D modeling
- Driving
- Mobile devices
- Accessibility...
Why is Visual Perception hard?
The gap between human perception and computational solution.
- Human: subjective perceptions
- Machine: numeric data
There are some challenges in this field:
- Images are 2D projections of the 3D world (Adelson and Pentland: The perception of shading and reflectance. Perception as Bayesian inference, 1996)
- Viewpoint Variation
- Deformation
- Occlusion
- Illumination
- Motion
- Perception vs. Measurement
- Local Ambiguities
- Intra Class Variation
- Number of Objection Categories
A brief history of CV
Credits:
Svetlana Lazebnik (UIUC): Computer Vision: Looking Back to Look Forward
Computer Vision: Looking Back to Look Forward (illinois.edu)
Steven Seitz (Univ. of Washington): 3D Computer Vision: Past, Present, and Future