End-to-end full-stack strength, Intel is using artificial intelligence to achieve the future
On May 27, 2017, Intel participated in the 2017 Global Machine Intelligence Summit (GMIS 2017), hosted by Machine Heart SYNCED. The conference focused on "Machine Intelligence Times", focusing on artificial intelligence, including Intel At home and abroad, the top experts gathered here, around the artificial intelligence to start a wonderful speech and intense discussion.
During the meeting, Dr. Liu Yinyin, Director of Intel's AIPG Data Science Department, delivered a keynote speech entitled "Evolving Artificial Intelligence, Accelerating the Model", said that in-depth learning promoted the progress of the field of artificial intelligence. Is a breakthrough in theory and practice, starting from the model, through the collection of data, training, based on artificial intelligence framework and computational power to solve the application problem, then the relevant experience feedback to the model to form a closed loop, Good model, to provide more efficient artificial intelligence solutions.
Four areas to see Intel's artificial intelligence use
From the past AI Day, to the present GMIS 2017, Intel has always stressed that in the field of artificial intelligence is to make a closed loop, in this ring, Intel to provide users with efficient artificial intelligence solutions, and is committed to more people to use Artificial intelligence technology, that is, AI democratization. However, by virtue of the existing artificial intelligence to achieve AI popularity is quite difficult, we all know, now the artificial intelligence does not achieve real intelligence. Such as AlphaGo can not be under the chess, in solving the problem, their intelligence range is also very narrow, can not be generalized, the breadth of the narrowest is the biggest obstacle to the popularity of AI, AI is the biggest dilemma, how can we from a narrow, specific areas Smart to more versatile intelligence? Intel gives the answer with the model.
On GMIS 2017, Liu Yanying shows how Intel can use a model to implement AI in different industries. Taking the most common image depth learning model as an example, image recognition can be used for agriculture to identify crop diseases. In the absence of this subdivision domain data set, Intel builds image recognition models for specific areas by adding layers and fine tuning to the trained generic image classifier to migrate learning methods. Similarly, in the automatic driving, oil and gas exploration, tumor detection and other tasks, the image model can also provide a large number of solutions. Today, Intel's Artificial Intelligence Division has applied the various models of the model to healthcare, retail, finance, transportation, government, industry, consumer goods and other industries.
In recent years, Intel deep plowing medical field, committed to the realization of precision medical through AI. On the basis of large data and artificial intelligence, Intel has worked with some of the US medical institutions to jointly carry out the Parkinson project and collaborative cancer cloud.
For Parkinson, Intel and the focus on Parkinson's disease research MichaeI J. Fox Fund, with AI expansion of human gene pool, in order to achieve difficult disease diagnosis and treatment.
In the area of cancer, Intel has partnered with the National Health Service System (NHS) and has signed a research project with the Alan Turing Institute for lung cancer cell identification. The project uses Intel Xeon processor, TensorFlow and other depth learning as the framework, Aimed at using AI to improve cancer detection efficiency.
In addition, Intel has partnered with PokitDok, a medical API company, to develop the first viable Dokchain medical block chain solution with open-source software for Intel's Hyperledger Sawtooth and the use of Intel chips for block transaction.
In the medical field, Intel to give full play to its cloud end of the technical advantages, is committed to AI through the reconstruction of medical models, to create high-quality health care services.
In the field of retail, Intel and McDonald's to work together to create a smart restaurant, and Wanda cooperation to help its fleeting wisdom business platform to provide more intelligent services, and Yi Jie Tong jointly launched smart POS cashier ... ... .., it is easy to see from the inventory Management, ordering, smart payment and traffic statistics and analysis, Intel in the retail field with a wealth of solutions, through the cloud to the end of the technological innovation and integration of hardware and software systems, Intel is committed to partners for smart restaurants, Intelligent department stores, smart convenience stores and other new retail model, in order to achieve the traditional retail industry to the wisdom of retail transformation, through the transformation to help customers achieve the value of data conversion.
In addition to smart retail, in response to retail, Intel also in the 2017 National Retail Electronics Show, launched a new Intel responsive retail platform (RRP). It is reported that RRP can achieve rapid tracking of inventory and user habits of timely feedback, and can provide users with personalized shopping experience, while Intel said the next five years will invest 100 million US dollars for the development and promotion of RRP.
In the field of transportation, Intel and Zhejiang Dahua to create the wisdom of traffic is Intel's artificial intelligence to use a typical, from the server, storage, large data and then to the cloud deployment, Intel to help Dahua to create the most advanced end-to-end traffic data Solution, and to provide Dahua based on Intel architecture embedded devices and hardware reference platform, in order to achieve a large number of video image content identification and traffic data analysis, to solve the Dahua long-term traffic video surveillance in the field of pain points. Recently, Intel is said, will Movidius's Myriad2 VPU technology for the designated Dahua camera, which will directly enhance the competitiveness of Dahua in the transport sector.
In addition, Intel in China has joined hands in Shenzhen Star Road, in the cross-road and many other auto manufacturers, jointly released based on the Intel Quark processor commercial vehicle network terminal, and successfully involved in vehicle information, automatic driving and other fields.
The Financial Sector
The field of finance is an important area of development for Intel's artificial intelligence. With its own technological innovation, Intel is making the "smart bank" a reality. In September, Intel launched a new financial counter trusted terminal in September, based on the Intel Skylake microprocessor architecture and adopted a new, Embedded operating system, to achieve a flexible development of different customer needs.
In addition to medical, retail, transportation and finance, similar applications also occur in other areas: in data security and privacy protection, Intel and Jingdong cooperation, in the last year on the eleven, the use of artificial intelligence accurate and rapid screening of illegal, In addition, Intel has also worked with a dedicated organization to help children in the United States, in order to protect the security of the city, the use of AI in the protection of urban security, the use of Intel in the computer vision and other technologies to combat crime and financial fraud, in addition, Technology to find missing children and abused children.....
Saving strength, Intel all-round to enhance their own AI strength.
It is not difficult to see that, through the end of the full-stack strength, Intel is gradually applied to various industries artificial intelligence. However, this is only the beginning of Intel's artificial intelligence path, as Intel's technology giant, is committed to the continuous improvement of artificial intelligence capabilities to promote the democratization of artificial intelligence, so that artificial intelligence to contribute to the whole society, so that everyone enjoy artificial intelligence To bring the well-being, this is where Intel's goal, in order to achieve this goal, Intel is constantly accumulating strength, all-round to enhance their own strength.
One, constantly extending the post-technology layout
In 2016, Intel acquired Nervana Systems for $ 400 million. As a leader in deep learning and neural network chips and software, Nervana has a fully optimized hardware and software stack for deep learning, and Intel is more competitive through technology integration Of the chip products, and in the silicon layer to achieve machine learning, with Nervana, Intel is the advantage of its own processor extends to the depth of learning.
In order to penetrate the machine vision, Intel in September 2016 acquisition of industry-leading visual processing chip maker Movidius, for computer vision, Movidius has a strong accumulation of technology, the acquisition of Intel to obtain a high expression, low-power Soc platform , Computer vision in the use of AI also accelerated. It is worth mentioning that, before that, Intel also acquired a computer vision algorithm specializing in the start-up company Itseez, no doubt, through these two acquisitions, Intel filled itself in the algorithm and the corresponding dedicated algorithm processor IP Of the lack of strength to be improved.
In the field of AI, Intel continued to extend the technology layout, in March this year, Intel to 15.3 billion US dollars acquisition of the world's leading unmanned program provider Mobileye, the acquisition is Intel's 48 years since the establishment of the second Large-scale acquisitions, Mobileye master camera, sensor chip and many other key technologies to achieve automatic driving, which makes Intel in the field of unmanned driving the core competitiveness of a substantial increase.
It is not difficult to see that Intel changed the low-key style in the past, in less than a year, with the strength of the rapid reinforcement of the AI field, through the acquisition coupled with their own product advantages, Intel can provide full stack strength to deal with end- Libraries and languages, frameworks, tools to applications, and Intel has all the assets and all the strength needed to provide end-to-end artificial intelligence solutions to the market.
Second, do more focus on AI's Intel
In order to focus more on artificial intelligence, in March this year, Intel announced that the company's all the artificial intelligence business together, set up a more focused on the AI artificial intelligence products division (AIPG), through the integration of the development of artificial intelligence Accumulated engineering, software, laboratory and other resources to create a more coherent unified AI ecological environment. Not only that, Intel has also set up an artificial intelligence application laboratory, specifically to promote the cutting edge of the work, explore new architecture and algorithms, the development of future artificial intelligence. Prior to that, Intel also worked with global agencies to provide developer training courses, set up by the AI field of experts in the artificial intelligence advisory committee for the next generation to provide artificial intelligence resources Nervana Artificial Intelligence Institute, but also with Google, including Of the industry's leading companies to form artificial intelligence alliance ... ... Recently, Intel has announced to join the AI cooperation organization, artificial intelligence in Intel's strategic position attention is evident.
For Intel, Intel is more important to promote the use of artificial intelligence in all walks of life than to promote the development of related technologies, that is, through AI democratization, so that more people enjoy the benefits of artificial intelligence, as technology Giant Intel, is now through its end to end full stack strength, change and promote the development of various industries.
Whether it is in the technical layout, or in other ways, Intel is all-round power for their own, through flexible end-to-end products and technical solutions for continuous innovation, Intel deep in the artificial intelligence revolution, and is committed to the use of artificial intelligence achievements The future, for artificial intelligence innovation and popularity, Intel has a firm determination, and look forward to re-casting in the AI field brilliant. Today, Intel is with its strong technical strength and business team to become a trusted leader in the field of artificial intelligence, with artificial intelligence to achieve a new future!
Resources link: http://www.kejilie.com/