The Current Situation and Future of AI Development from the Perspective of Patent

About Artificial Intelligence (AI) The research has been carried out in full swing all over the world, and the development of different regions is different. At present, there are a lot of research on AI from various aspects.Technology is discussed, and patented technology is an important indicator to reflect the real technical strength. By analyzing the application data of patents related to AI, we can systematically and comprehensively analyze the development status and future development direction of AI, including the trend characteristics of AI development, regional characteristics, the most innovative research and development institutions, major technology exporting countries and regions, etc.

   By analyzing the patent application data of different countries and regions, the following conclusions can be drawn:

1.China Surpasses Traditional Science and Technology Power to Become AICountries with the highest number of patent applications. After nearly two years of rapid development, China's AIThe number of patent applications surpassed the United States and Japan, ranking first in the world.

Currently AITop 10 countries by number of patent applications/The regions are China, the United States, Japan, South Korea, Europe, Germany, Australia, Taiwan, Canada and India. The top two, China and the United States, are far ahead of the rest of the world.

Top 10 countries and regions in number of patent applications

2.In the past two years, the number of patent applications has maintained a strong upward trend, except in China,At the same time, the number of applications in the United States, Japan and Europe grew slowly, especially in 2016.The number even began to fall around the year. Although possible and partially 2017The patents filed in the year were not disclosed, but the overall decline in the number of AI patent applications seems to have been difficult to reverse.

3.Baidu and IBMChinese and American AI respectivelyResearch institutions with the most patent applications. Although Baidu started late in this field, it has a strong momentum and has come from behind, far surpassing other applicants.

In addition, the Chinese Academy of Sciences and Tencent also performed well. Although the two foreign-funded companies Microsoft and Samsung once had an advantage in the number of applications, they have shown a trend of slowing growth in the past two years.

China's main AIRanking of Patent Application Institutions

4.American AIThe company with the most patent applications is IBM.. Google, which has been vigorously researching and developing AI in recent years, is not ideal in the number of AI patent applications in the United States, ranking only third, and its number of applications (1885) is far from IBM(4322), which ranks first. Only about half of the top companies in the United States are American companies, and the other seats are occupied by Japan and South Korea. On the contrary, the layout of American enterprises in Japan and South Korea is not much.

5.The world's most applied AIPatents are related to natural language processing. The most representative branches in the field of artificial intelligence-machine learning and basic algorithms, intelligent search and intelligent recommend, speech recognition, natural language processing, autonomous driving, and computer vision and image recognition. The number of global and domestic patent applications is shown in the following table.

Various technical branches in 2010After the rapid growth, especially machine learning and basic algorithms, in the 2014.After the year, it was a standout, almost in a straight line, growing much faster than the other branches. Autonomous driving, which started late, grew faster than other branches after 2014, but declined slightly after 2017. The number of applications from various technical branches in the United States as a whole showed a tortuous upward trend. After several years of rapid development, the growth rate began to decline after 2014. The development trend of each technology branch in Europe is similar to that in the United States.

Number of patent applications for each technology branch of AI

6.Domestic R & D institutions have their own strengths, and they have advantages in patent applications for different technical branches. In the direction of machine learning and basic algorithms, research institutes and universities occupy more seats, ranking in the top 15.Of the applicants, 10Bit. The remaining five are Microsoft, Baidu, Tencent, Alibaba and State Grid. It can be seen that the current domestic research on machine learning and basic algorithms is still in the laboratory stage, although the number of patents is large, there is still a certain gap from commercial applications. In the direction of intelligent search and intelligent recommend, in contrast to machine learning and basic algorithms, patents are mainly concentrated in large Internet companies and intelligent terminal manufacturers.SimultaneouslyVoice recognition direction front 15The number of applicants is mainly enterprises, and there is only one scientific research institution in the Chinese Academy of Sciences. Among them, Baidu emerged from the siege of a number of foreign-funded enterprises and topped the list.AndIn the natural language processing technology branch of the distribution is relatively average, enterprises and research institutes and applicants respectively occupy half of the country. The top three Baidu, the Chinese Academy of Sciences and Microsoft's overall application volume is more prominent. IBMThe ranking has reached a relatively high position in this field. It is worth noting that the Chinese Academy of Sciences and Zhejiang University are among the best in many fields. At present, many enterprises have entered the experimental stage in the direction of automatic driving.In the direction of automatic driving, but these enterprises are mainly Ford, Toyota as the representative of the old foreign car manufacturers. Baidu has invested heavily in self-driving in recent years to get the top spot, while other domestic automakers have performed mediocre.FinallyIn the direction of computer vision and image recognition, the distribution is also relatively even, with roughly half of enterprise applicants and half of scientific research institution applicants. Due to the increasing demand for image processing of intelligent terminals, it has promoted the R & D investment of intelligent terminal manufacturers in this field, making enterprises have an advantage in the number of patent applications.

future development trend

2010Years later, AIIt has ushered in a period of rapid technological growth, with the largest number of patent applications in China, which still maintains a strong growth trend. The United States, after a period of rapid growth, has slowed down, while Japan has stagnated.Although China has a large number of applications as a whole,However, the number of overseas applications is seriously low, that is to say, the number of high-quality and high-value patent applications is relatively small. Our AIThe development of technology is still not optimistic compared to foreign countries.Also AIThe number of applications for computer vision and image recognition in all branches has increased significantly in recent years, and will become AI in the future.A huge growth point in the industry.

Summary of domestic AI patents

Generally speaking, the domestic technology development in the field of artificial intelligence is thriving and is in the stage of rapid technological growth. With the rapid growth of patent applications by domestic enterprises, R & D investment is bound to be huge. In order to do a better job of patent layout and take the lead in this technological innovation, enterprises can make breakthroughs in the following aspects in the technological research and development in the field of artificial intelligence.

Even in AI.The number of patent applications in the field is huge, but most of them are the achievements of scientific research institutes, that is, most technologies are still in the laboratory stage, there is still a certain distance from commercial application, and there is a certain gap between technology and commercial use and foreign countries. In the future development, we should not only pay attention to high-quality patents and overseas patent applications, but also speed up the transformation of patent achievements and speed up AI.The pace of technology commercialization, in the commercial competition to obtain a real technological advantage.

Article Source: http://www.sohu.com/a/275291151_473283