Thema der Dissertation: Data-driven Disease Assessment from time-resolved Fluorescence Optical Imaging Thema der Disputation: Generative Adversarial Networks
Abstract: In biological and medical imaging, data acquisition and data annotation are expensive. However, for methods such as Convolutional Neural Networks (CNN), large annotated datasets are required for good results. Generative Adversarial Networks (GANs) can generate images, that resemble the data that they have been trained on . In contrast to other generative approaches,GANs do not require the optimization of a loss function, that quantifies the reconstruction error...... The full abstract can be found at the following link.
Zeit & Ort
13.07.2021 | 11:30