Suche

zur Hauptseite

Institute for Advanced Simulation (IAS)

Navigation und Service


IAS-Seminar "Image Restoration with Generative Adversarial Networks (GANs)"

Anfang
20.02.2020 14:00 Uhr
Ende
20.02.2020 15:00 Uhr
Veranstaltungsort
Jülich Supercomputing Centre, Rotunde, Geb. 16.4, R. 301
Referentin:
Tetiana Martyniuk, Machine Learning Lab, Lemberg University, Ukraine
Abstract:

Image restoration is one of the key computer vision tasks that aims to transform the corrupted image in such a way that it becomes as close to the original target as possible in the general visual appearance and perceptual quality. It is also used as a preprocessing step for other computer vision (CV) algorithms, as for instance many pretrained models require non-degraded image data.

Recently, Generative Adversarial Networks (GANs) led to tremendous progress in this field, treating the task as image-to-image translation conditioning on the input. In this talk, I will give an overview of existing methods, present the details and results of my thesis and the latest joint ICCV paper on image deblurring, and discuss the challenges of evaluation and measurability of perceptual quality.

Zeit:
Donnerstag, 20. Februar 2020, 14:00 Uhr
Ort:
Jülich Supercomputing Centre, Rotunde, Geb. 16.4, R. 301

Julia Sidorova wurde eingeladen von Dr. Jenia Jitsev (JSC).


Servicemenü

Homepage