ETH Zurich – “Low-end and compact mobile cameras demonstrate limited photo quality mainly due to space, hardware and budget constraints. In this work, we propose a deep learning solution that translates photos taken by cameras with limited capabilities into DSLR-quality photos automatically. We tackle this problem by introducing a weakly supervised photo enhancer (WESPE) – a novel image-to-image GAN-based architecture. The proposed model is trained by weakly supervised learning: unlike previous works, there is no need for strong supervision in the form of a large annotated dataset of aligned original/enhanced photo pairs. The sole requirement is two distinct datasets: one from the source camera, and one composed of arbitrary high-quality images – the visual content they exhibit may be unrelated. Hence, our solution is repeatable for any camera: collecting the data and training can be achieved in a couple of hours. Our experiments on the DPED, Kitti and Cityscapes datasets as well as on photos from several generations of smartphones demonstrate that WESPE produces comparable qualitative results with state-of-the-art strongly supervised methods.”

“In this work, we presented WESPE – a weakly supervised so-
lution for the image quality enhancement problem. In contrast
to previously proposed approaches that required strong supervi-
sion in the form of aligned source-target training image pairs,
this method is free of this limitation. That is, it is trained to map
low-quality photos into the domain of high-quality photos with-
out requiring any correspondence between them: only two sep-
arate photo collections representing these domains are needed.
To solve the problem, we proposed a transitive architecture that
is based on Generative Adversarial Networks and loss functions
designed for high-quality image quality assessment. The method
was validated on several publicly available datasets with differ-
ent camera types. Our experiments reveal that WESPE demon-
strates the performance comparable or surpassing the traditional
enhancers, and gets close or competes with the current state of
the art supervised methods, while relaxing the need of supervi-
sion thus avoiding tedious creation of pixel-aligned datasets.”

PDF

TRY IT

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.