FakeNews: GAN-based generation of realistic 3D volumetric data -- A systematic review and taxonomy

arXiv:2207.01390v2 Announce Type: replace-cross
Abstract: With the massive proliferation of data-driven algorithms, such as deep learning-based approaches, the availability of high-quality data is of great interest. Volumetric data is very important in medicine, as it ranges from disease diagnoses to therapy monitoring. When the dataset is sufficient, models can be trained to help doctors with these tasks. Unfortunately, there are scenarios where large amounts of data is unavailable. For example, rare diseases and privacy issues can lead to restricted data …

abstract algorithms arxiv availability cs.cv cs.db cs.lg data data-driven dataset deep learning disease fakenews gan massive medicine monitoring quality quality data review taxonomy therapy type