Neural networks are a powerful tool used in many fields of science and technology. They represent a novel approach to data processing, replacing, at least on routine basis, a painstaking human work, which opens new horizons in the reproducibility, quality and quantity of scientific results. Neural networks have been recently applied to tasks such as pattern recognition and enhancing the quality or improving the angular resolution of astrophysical images. The proposed research is focused on the development and application of neural networks in observational extra-galactic astronomy. It includes the detection and classification of objects and the determination of their empirical, statistical and physical characteristics. The student should have a background in extragalactic astrophysics and should be familiar with recent neural network frameworks, as well as standard data processing techniques.