Main Content
Computer Vision I (Lecture & Exercise)
The lecture Computer Vision I provides a comprehensive introduction to how computers can represent, analyze, and understand images and videos. It covers classical image-processing foundations as well as modern machine-learning and deep-learning methods for visual recognition and understanding.
The course starts with the fundamentals of digital images, color spaces, linear filtering, and transformations. It then introduces machine learning for vision, covering gradient-based learning, generalization, neural networks, convolutional neural networks, and transformer models. Building on these foundations, it explores representation learning, generative models, pre-training, and transfer learning. The final part of the lecture applies these techniques to central vision tasks, including motion analysis, depth estimation, segmentation, and vision–language problems.
The lecture will teach both the theoretical principles and practical methods behind modern computer-vision systems, from image filtering to large-scale neural models, reflecting the field’s progression from image processing and feature extraction to deep architectures, generative models, and multimodal vision–language applications.