In an age of selfies, photoshopped Facebooks and Instagram, you can’t resist the continual need to fix all those grainy or pixelated images in your photo library. Say hello to Nvidia’s latest AI (Artificial Intelligence) technology, which can not only automatically turn grainy pictures into pretty ones but also remove text and watermarks.
Developed by researchers from NVIDIA, Aalto University, and MIT, the work is being presented at the International Conference on Machine Learning in Stockholm, Sweden this week.
It is worth mentioning here that AI is not being used to fix grainy photos for the first time here. But the difference lies in the fact that in previous iterations, the AI needed to look at before-and-after photos with both corrupted and optimal examples to fix it. However, the latest AI technology can fix pixelated photos without seeing a cleaner version of the target pictures.
“It is possible to learn to restore signals without ever observing clean ones, at performance sometimes exceeding training using clean exemplars,” the researchers said in their paper. “[The neural network] is on par with state-of-the-art methods that make use of clean examples — using precisely the same training methodology, and often without appreciable drawbacks in training time or performance.”
Using NVIDIA Tesla P100 GPUs with the cuDNN-accelerated TensorFlow deep learning framework, the team trained their system on 50,000 images in the ImageNet validation set.
To test the system, the team validated the neural network on three different datasets.
The method can even be used to enhance MRI images, perhaps paving the way to drastically improve medical imaging.
“There are several real-world situations where obtaining clean training data is difficult: low-light photography (e.g., astronomical imaging), physically-based rendering, and magnetic resonance imaging,” the team said. “Our proof-of-concept demonstrations point the way to significant potential benefits in these applications by removing the need for potentially strenuous collection of clean data. Of course, there is no free lunch – we cannot learn to pick up features that are not there in the input data – but this applies equally to training with clean targets.”