AI never ceases to amaze researchers and developers in the case of showcasing its intelligence and abilities to render extraordinary results thanks to machine learning capabilities. Recently, Google developed an AI that is capable of capturing the perfect photos out of just simple street view panoramas that are so good, that 40 percent of the photos were deemed to be either pro level or semi-pro level photography by experts in the field.
The AI uses something called Creatism system that has not yet been offered publicly by Google just yet. Apart from simply cropping shots from panoramas taken by Street View, the AI will enhance the images and adjust tones, saturation and other features, as well as add HDR and lighting effects to make the photos look pro-level. Even professional photographers claimed that the pictures were in-fact semi-pro level photography.
Reports mentions that the AI studied over 40,000 panorama shots in areas like the Alps, Banff and Jasper National Parks in Canada, Big Sur in California and Yellowstone National Park. The photos that the researchers had trained the AI in learning were all professional quality rather than before and after shots, bringing to light that even artificial intelligence can mimic the work of a professional quality photographer. This will enable developers to create programs and systems that will allow for users to create similar high-quality images in the real world as well.
About 40 percent of the outputted photos edited by the AI were considered by professional photographers to be either pro or semi-pro in quality. While these are positive results, it still means that not all the photos that have been captured by the AI matched the same consistency, owing to the machine learning process that improves the AI’s results over time. Considering that Google Maps has over one billion monthly active users, this would help in generating better shots for viewers to see, and would greatly help in amplifying the experience of using the app, should users want to explore certain scenic points of areas of a region.
Apart from efforts in the Creatism System, Alphabet is also focused on other AI projects, which include DeepDream image enhancement software as well as the NSynth synthesizer. Perhaps when these systems have reached a certain level of maturity and development, they will be more widely available to the public for simple use in creativity apps. This is just another application that AI has shown that it can excel in.