J. Hartz, L. Hotz, B. Neumann, and K. Terzić (Germany)
Computer Vision, Image Understanding, Machine Learning, Ontologies
In this paper, we investigate automatic model learning for the interpretation of complex scenes with structured objects. We present a learning, interpretation, and evaluation cycle for processing such scenes. By including learning and interpretation in one framework, an evaluation and feedback learning is enabled that takes interpretation challenges like context and combination of diverse types of structured objects into account. The framework is tested with the interpretation of terrestrial images of man-made structures.
Important Links:
Go Back
