Apple is finally no longer obscured and released its first artificial intelligence research report. According to industry insiders, this will be of great benefit to Apple's future promotion of its artificial intelligence applications.
Apple said earlier this month that it will publish its own artificial intelligence research report. In less than a month, Apple fulfilled its promise. Recently, Apple released the first academic paper on artificial intelligence.
This report describes a new technique for training image recognition capabilities of an algorithm by computer-generated images rather than real images.
Apple said in the report that in machine learning research, using synthetic images (such as from a video game) to train neural networks is more effective than using real images. Because the composite image data has been tagged and annotated, real image data requires a lot of effort to mark everything the computer sees, such as a tree, a dog, or a bicycle.
Of course, the use of composite images also has certain drawbacks, resulting in an algorithm that understands something different from the real-world scene.
Apple said in the report: "Sometimes, the composite image data is not real enough, so that the neural network can only understand the details presented in the composite image, but the understanding of the real image is insufficient."
In order to solve this problem and improve the training effect of synthetic image data, Apple researchers have introduced a "simulated + unsupervised" learning method to improve the realism of the simulated image.
Apple researchers use a modified new machine learning technique called the Generate Against Network (GAN) that lets two neural networks fight each other to produce more realistic images.
The first author of this report is Apple researcher Ashish Shrivastava, who holds a Ph.D. in computer vision from the University of Maryland, College Park.
According to industry insiders, for Apple, the publication of its first artificial intelligence research report is also a big step forward. Over the years, Apple has been keeping a tight eye on its research in the field of artificial intelligence, which has been criticized by the artificial intelligence research community. At the same time, this also affected Apple's recruitment of artificial intelligence talent.
In addition, this disclosure of its own artificial intelligence research will also help to popularize its own artificial intelligence software in the future. Today, artificial intelligence software is being applied to almost any application, such as the camera function of the iPhone 7, and various Internet services.
Silicon windows are ideal for IR applications with good transmittance from 1.2 to 7.0µm. Silicon has a low 2.329 g/cm3 density which is half the density of Germanium, making it ideal for weight sensitive applications.
Silicon is deal for 1.2 to 7 µm NIR applications
Silicon is broadband 3 to 12 µm anti-reflection coating
Silicon is ideal for weight sensitive applications
Si Windows And Lens,Si Crytal Lens,Silicon Windows,Dlc Coating Silicon Windows
Zoolied Inc. , https://www.zoolied.com