Visual Recognition and Analysis / Objekterkennung

This course introduces to advanced visual recognition techniques, which are important to analyze image and sensor data automatically.

• Brief summary of basic machine learning techniques (nearest neighbour, SVM)
• Local and high-level global descriptors (histograms of gradient orientations)
• Image categorization with supervised and unsupervised methods
• Localization with sliding window methods (Haar-like features, boosting, random forests, feature maps and convolution) 
• Semantic segmentation (convolutional neural networks, available toolboxes, exercises based on medical data and counting tasks)