Xiaomi the development of artificial intelligence DeepExposure to restore details of the image high exposure


The cameras in our smart phones today, rely on the sensor to capture images and processor for image processing after that. Now, the company introduced the Xiaomi method aims to solve a common problem of pixels is small, a dynamic range limited, through the use of technology and artificial intelligence to fix exposure in photos.

But not the entire image, instead the image is divided into image sub and adjust the exposure separately. Then it is integrating the different parts to form the final image. Take this image for example, the buildings white, and the clouds white and the sky high exposure to white as well. Artificial intelligence developed by company Xiaomi, which took the name of DeepExposure, and did a wonderful job in the back details.


The training of artificial intelligence, the use of the engineers in the company Xiaomi images from the database of the MIT-Adobe FiveK. Contain the RAW images unedited in addition to the same images that have been revised by five experts note that the engineers in the Chinese company used the 3000 images, with a choice of images that have been retouch by distinguished experts.

The technology works on low resolution images, the primary goal is to invent the best indicators to filter the images classic. The idea to include this technique in Adobe Lightroom. This simplifies the process of learning, but should also lead to speed up image processing.



The post Xiaomi the development of artificial intelligence DeepExposure to retrieve details of the high sound exposure appeared first on electronic.

Leave a Reply

Your email address will not be published. Required fields are marked *