One of the biggest issues facing consumer-level drones today is the image quality. Whether you’re talking low light shots or shooting dark objects, you will face the issue of a grainy image. This is soon to change thanks to a collaboration between Nvidia, MIT and Aalto University.
Nvidia has built something they are calling Noise2Noise, which is AI software that can clear up images from drones and more. Noise2Noise is a great use of a neural network, an image database, and deep learning. Let’s see what it’s all about.
[related_articles title=”Related Articles”][/related_articles]
What is Nvidia Noise2Noise?
We all know the feeling of taking an amazing photo, going to share or edit it only to realize that it has been ruined due to graininess caused by low light in the image. It’s frustrating. Luckily, Nvidia is working to fix this for you. With Noise2Noise, Nvidia is utilizing a neural network, a massive image database, and deep learning to repair images, and the results are fairly impressive.
Noise2Noise takes an image, analyzes it and figures out what adjustments need to be made. It will then take the data it collected from the image and fill in gaps between already placed colors. It will do this by looking at the colors around the white space to ‘guess’ what color would be the best fit to go in the gap.
Right now Noise2Noise is used to enhance grainy images that are affected by low light. This is a great use for sure, but there are many more possibilities where Nvidia’s AI can help out.
Future uses for this type of AI include; low-quality photos, old photos, damaged photos and even the creation of new photos. The uses of this type of AI are endless, as time progresses new uses will show through.
Nvidia is no doubt one of the largest companies that are currently taking on such a massive issue that most people have faced before. Do you think Nvidia will release this AI to the world? Share your thoughts with us in the comments below or head over to our social media accounts.