[Global Technology Reporter Xinyue] On the evening of March 21, Cheetah Mobile announced the third quarter and 2016 full-year financial statements as of December 31, 2016. In the fourth quarter, total revenue was 1.267 billion yuan, overseas revenues increased by 31.3% year-on-year, mobile revenues accounted for more than 80%, and total revenue in 2016 was 45.647 billion yuan, a year-on-year increase of 21%. With the help of artificial intelligence technology, Cheetah Mobile has created historical highs in many important indicators such as operating income, mobile income, and overseas income. Cheetah’s awareness of artificial intelligence and deep learning has already taken a leading position in the industry. Coupled with more than 600 million monthly active users and big data, Cheetah is able to achieve corner crossing in artificial intelligence and deep learning. Artificial intelligence reshapes the entire industry and reshapes the whole thinking method of cheetahs. Deep learning is a major technological revolution. Its essence is to make the algorithm be normalized. The essence of data-driven is to make the algorithm be normalized. The essence of data-driven is to make the algorithm be normalized, and the data-driven budgeting ability becomes very important. Artificial Intelligence Drives Content Platform In 2017, Cheetah is making an all-out effort to upgrade from a utility software company to a content platform company driven by artificial intelligence. Cheetah will use artificial intelligence to achieve the new mission of the next decade - Make the World Smarter. According to Fu Sheng, CEO of Cheetah Mobile, “AI is a basic tool and a technical revolution in the software industry. Many of the inherent algorithms in the software field have lost their usefulness because AI can be implemented in a simpler manner and may take a long time before. Unavailable technology. So I think this is a huge opportunity and a key direction. Cheetah will first seek AI technology to be used in many of our products. For example, we will use a deep learning approach to make content recommendation engines." Artificial intelligence and product integration deeper Cheetah is using artificial intelligence more and more in Live.me. Currently, artificial intelligence technology involves businesses such as anti-spam and personalized recommendations. Specifically, first, recognition of erotic information is performed through background technology detection, and an official logo can also be identified and advertisement detection can be performed. Secondly, it is possible to classify the anchor by gender, race, type (Uncle, fresh meat, male, female, etc.), and to perform face monitoring (such as similarity, whether it is a celebrity face), age detection, and value. Classification, based on this classification, we can more accurately understand user preferences and provide personalized content recommendations for users. In addition, various dynamic face maps can also be implemented. Cheetah Mobile has conducted an in-depth revision of News Republic, AI+Cheetah+News Republic (adopts interest information application). It uses advanced learning models to replace News Republic’s original expert model to increase the user’s conversion rate so that each user can pick up the phone. The news is unique, which greatly enhances the user experience, user activity, and usage time. As a distribution system of artificial intelligence, DNN understands the user and understands the content more deeply. It pushes the right content to the people who like it and truly realizes thousands of people. Cheetah also uses artificial intelligence in PhotoGrid. In terms of image recognition, classification attributes such as food, scenery, portraits, and the like can be automatically identified based on pictures uploaded by the user. On the recommendation of personalization, recommendation of matching content can be performed according to the user's attributes, interests, and hobbies; In terms of processing, through the application of artificial intelligence, dynamic effects can be increased, traditional photographs can be turned into dynamic and oil painting effects, facial effects, and dynamic filters.