Nvidia's new AI program greatly benefits low-light photos by eliminating unwanted grains

Nvidia Graphics Maker has developed a new artificial intelligence program that will make granular photos a thing of the past.

Nvidia worked with researchers at Aalto University and MIT to find an automatic mechanical solution to make granular photos as clear as possible. Photos taken in low light conditions are usually full of noise and artifacts. Using this deep learning approach, the technology will be able to fix granular or pixelated images by simply referring to a prototype of degraded photos.

Further research has been done in this similar area. including the Facebook AI model designed to replace a flashing eye in photos. What makes the Nvidia model different is that it only needs two input photos with noise or grain for the job to be done.

The team formed the model on 50,000 images using Nvidia's Tesla P100 GPUs. According to the research team, their model can automatically enhance photos by getting rid of artifacts, noise and grain without having to see what a clear photo looks like.

It is possible to learn to restore signals without ever observing them to performance sometimes exceeding the training using clean copies. [The neural network] is up to the cutting edge methods that use clean examples – using precisely the same training methodology, and often without appreciable disadvantages in terms of training time or performance. "

Far from what it is At this point, the model can be used in areas where getting clean driving data is difficult, especially low-light photography, such as astronomical imaging, MRI images, magnetic resonance imaging and physical rendering

Concept demonstrations pave the way for significant potential benefits in these applications by eliminating the need for A potentially laborious collection of clean data, "said the team.

But for this to be fully effective, the program needs input data Here are some examples of Nvidia's photos: