Lectures and assignments

Lectures

Assignments - Final project (or computer vision "hack-week")

The framework mostly used in class will be python and openCV, but students are free to use any other language/framework (recommended to use openCV libraries) in their final project.

Starting from the demo and exercises code given in each lecture, use two (or more) of these "ingredients" to build your final project

• Deep features
• BoW
• Image segmentation
• Face detection
• Person detection
• Tracking
• Action classification

Develop your project towards your application of choice, using one of these datasets:

• People datasets ("static" images):
  • Urban Tribes
  • Faces
  • Clothing
• "Ego-centric" (first person perspective) data (videos):
  • RGB-d sequence from a robot mounted camera
  • Soccer robots camera

For additional options (such as live processing of soccer robot images) discuss with the professor.

More details about final project presentation TBA