China's Race for Artificial Intelligence (AI) Technology

[cirr]

Tencent to advance into new AI areas

2018-03-03 10:47 China Daily Editor: Mo Hong'e

Visitors look at Tencent's visible big data exchange platform at the 16th China Internet Conference in Beijing on July 12, 2017. (Photo/for China Daily)

Tencent Holdings Ltd is looking to deepen its artificial intelligence push into transportation solutions, security and protection, as well as voice recognition, realms that were once seen as strongholds of other tech majors.

The endeavor to advance into new areas builds upon the company's initial success in developing AI for medical diagnosis, a mission designated to Tencent by the country's top industry and internet regulator, according to Liu Yongsheng, head of Tencent AI Lab.

"For the purpose of better city management, there's plenty of space for different companies to leverage and deploy their AI capabilities. Every player is just taking a tiny slice of the pie," he told China Daily on Friday.

Ideally, algorithms and big data can be combined to provide live traffic predictions, so that safety precaution measures and contingency plans can be introduced in advance, he said.

This is an area where arch rival Alibaba Holding Group Ltd has gained an early foothold. It pioneered a smart traffic management system called City Brain in Hangzhou, where average traffic speed was reported to have increased 15 percent since its launch in September 2016. Early this year, the firm exported a similar mechanism to Malaysia, the first of its kind overseas.

"Nevertheless, there are just so many facets in urban planning that companies can dig deep into," Liu said. "We don't necessarily have to compete with each other and can all be part of city planning in the long run."

Voice recognition is another area on which Tencent is placing bets. An indigenous group is working to optimize accuracy from voice conversion to machine translation, functions that are embedded in WeChat, the ubiquitous messaging app that has nearly 1 billion users.

"We are developing our own team and own technologies," Liu said, when asked if they are banking on technologies of iFlyTek, a domestic voice recognition specialist.

China has upheld the development of AI as a national strategy and has recruited Tencent, Alibaba, iFlyTek and Baidu Inc to an "AI national team" to have them each focus on one specific field.

While Tencent has been tasked to use its computer vision for medical diagnosis, Baidu will specialize in autonomous driving, Alibaba in smart city and iFlyTek in voice intelligence.

According to Liu, Tencent will further expand the use of its AI-backed cancer-detection product Miying to a wider array of diseases.

Currently it is used in reading CT scans and screening of four types of cancers including lung cancer, the most common cancer in China, in around 60 hospitals across 20 provinces.

"We are anticipating an even bigger role for Miying this year, by having AI assist human doctors in their daily diagnosis," he said.

http://www.ecns.cn/2018/03-03/294403.shtml

 

[cirr]

China's AI-general practitioner system starts hospital trial

2018-03-05 16:21 Xinhua Editor: Gu Liping

An AI-general practitioner system developed by a Chinese tech firm has started its "internship" in a community hospital in east China's Anhui Province.

The system called "AI doctor assistant" can listen to doctors diagnosing inquiries with patients and automatically produce e-documents for patient case reports.

Hu Jingyun, director of the Shuanggang Community Health Service Center in Luyang District of Hefei, the provincial capital, said the AI assistant could quickly review patient case histories and suggest prescriptions based on data in similar cases.

The system was developed by Shenzhen-listed iFlytek in partnership with Tsinghua University.

The company's medical robot passed China's national medical license examination with a high score in 2017. It was the world's first robot to pass a national medical license examination.

The company said it had optimized the robot technology in the development of the "AI doctor assistant," equipping it with intelligent voice functions and self-learning abilities.

Currently, the system's diagnosing reports and prescriptions still need the signatures of doctors for approval.

Ke Manxue, a health official in Luyang District, said the district would use the "AI doctor assistant" to improve grassroots medical services, as it could ease the workload of general practitioners in outpatient visits.

http://www.ecns.cn/2018/03-05/294639.shtml

 

[cirr]

MIT, China’s SenseTime will partner on AI research

By Fred Bazzoli

Published March 05 2018, 4:28pm EST

The Massachusetts Institute of Technology is partnering with a Chinese artificial intelligence company as it begins to research breakthroughs that can impact healthcare, imaging and other technologies to improve the human condition.

MIT says the alliance with Beijing-based SenseTime is the first in its MIT Intelligence Quest, launched in February. The technology school’s initiative seeks to leverage its strengths in brain and cognitive science and computer science to advance research into human and machine intelligence.

In particular, the MIT-SenseTime Alliance on Artificial Intelligence, as it is formally called, expects to study areas such as medical imaging, robotics, human intelligence-inspired algorithms and computer vision.

MIT executives say an essential element of its MIT Intelligent Quest initiative is forging connections with innovative companies and individuals who share MIT’s passion for work in intelligence.

SenseTime was founded by MIT alumnus Xiao-ou Tang and specializes in computer vision and deep learning technologies. The company says SenseTime has worked with more than 400 leading customers and partners to solve real-world problems. Among its current projects are intelligent medical treatment, deep learning hardware and optimization and autonomous driving.

MIT operates a Computer Science and Artificial Intelligence Laboratory, MIT Media Lab, Department of Brain and Cognitive Sciences, Center for Brains, Minds and Machines and the MIT Institute for Data, Systems and Society.

https://www.healthdatamanagement.com/news/mit-chinas-sensetime-will-partner-on-ai-research

 

[TaiShang]

Tencent to advance into new AI areas
By He Wei
China Daily, March 3, 2018

Tencent Holdings Ltd is looking to deepen its artificial intelligence push into transportation solutions, security and protection, as well as voice recognition, realms that were once seen as strongholds of other tech majors.


The endeavor to advance into new areas builds upon the company's initial success in developing AI for medical diagnosis, a mission designated to Tencent by the country's top industry and internet regulator, according to Liu Yongsheng, head of Tencent AI Lab.


"For the purpose of better city management, there's plenty of space for different companies to leverage and deploy their AI capabilities. Every player is just taking a tiny slice of the pie," he told China Daily on Friday.


Ideally, algorithms and big data can be combined to provide live traffic predictions, so that safety precaution measures and contingency plans can be introduced in advance, he said.


This is an area where arch rival Alibaba Holding Group Ltd has gained an early foothold. It pioneered a smart traffic management system called City Brain in Hangzhou, where average traffic speed was reported to have increased 15 percent since its launch in September 2016. Early this year, the firm exported a similar mechanism to Malaysia, the first of its kind overseas.


"Nevertheless, there are just so many facets in urban planning that companies can dig deep into," Liu said. "We don't necessarily have to compete with each other and can all be part of city planning in the long run."


Voice recognition is another area on which Tencent is placing bets. An indigenous group is working to optimize accuracy from voice conversion to machine translation, functions that are embedded in WeChat, the ubiquitous messaging app that has nearly 1 billion users.


"We are developing our own team and own technologies," Liu said, when asked if they are banking on technologies of iFlyTek, a domestic voice recognition specialist.


China has upheld the development of AI as a national strategy and has recruited Tencent, Alibaba, iFlyTek and Baidu Inc to an "AI national team" to have them each focus on one specific field.


While Tencent has been tasked to use its computer vision for medical diagnosis, Baidu will specialize in autonomous driving, Alibaba in smart city and iFlyTek in voice intelligence.


According to Liu, Tencent will further expand the use of its AI-backed cancer-detection product Miying to a wider array of diseases.


Currently it is used in reading CT scans and screening of four types of cancers including lung cancer, the most common cancer in China, in around 60 hospitals across 20 provinces.


"We are anticipating an even bigger role for Miying this year, by having AI assist human doctors in their daily diagnosis," he said.

 

[JSCh]

Inside the Chinese lab that plans to rewire the world with AI - MIT Technology Review

 

[JSCh]

Upgraded Deep Voice can mimic any voice in mere seconds
March 6, 2018 by Bob Yirka, Tech Xplore

​

Speaker adaptation and speaker encoding approaches for training, cloning and audio generation. Credit: arXiv:1802.06006 [cs.CL]

Via whitepaper which they have uploaded to the arXiv preprint server, a team at Baidu (China's answer to Google) has announced an upgrade to their text-to-speech application called Deep Voice. Now, instead of taking a half-hour or longer to analyze a person's voice and replicate it, the system can do it in less than a minute. The neural-network based system is part of an effort by the team at Baidu to make machines sound more like humans when they "speak" to us.

There are two parts to the system. The first involves recording voice samples to allow the system to learn what the subject's voice sounds like. The second part reads user-defined text aloud in the voice of the subject.

Several groups have been working on projects aimed at replicating the sound of an individual person's voice, ostensibly to allow robot assistants to sound like actual human assistants. Thus, a program that converts text into words that sounds like you, your neighbor, Donald Trump or the Queen of England is not expected to offer much in the way of an end product—though Baidu does suggest it could be used by people who have lost the use of their voice. Instead, it is meant as a stepping stone to greater things. The new system, the team reports, works optimally when given 100 five-second voice samples. It can also manipulate a voice, allowing people to hear how they might sound, for example, with a British accent, or as someone of the opposite gender. It is also getting better at mimicking voices, and is now able to fool voice recognition software 95 percent of the time—and a human test gave the system an average rating of 3.16 out of 4.

But, as many in the press have noted, the technology could cause problems. Taped interrogations by police could become useless if anyone with a smartphone could generate the same conversation. There is also the problem of identity theft. If a thief can steal your data and your voice, you might never get it back. Or consider political operatives releasing fake recordings of politicians having conversations that could sway an election.

More information: — Baidu Research blog: research.baidu.com/neural-voice-cloning-samples/

— Samples: audiodemos.github.io/

— Neural Voice Cloning with a Few Samples, arXiv:1802.06006 [cs.CL] arxiv.org/abs/1802.06006

Abstract
Voice cloning is a highly desired feature for personalized speech interfaces. Neural network based speech synthesis has been shown to generate high quality speech for a large number of speakers. In this paper, we introduce a neural voice cloning system that takes a few audio samples as input. We study two approaches: speaker adaptation and speaker encoding. Speaker adaptation is based on fine-tuning a multi-speaker generative model with a few cloning samples. Speaker encoding is based on training a separate model to directly infer a new speaker embedding from cloning audios and to be used with a multi-speaker generative model. In terms of naturalness of the speech and its similarity to original speaker, both approaches can achieve good performance, even with very few cloning audios. While speaker adaptation can achieve better naturalness and similarity, the cloning time or required memory for the speaker encoding approach is significantly less, making it favorable for low-resource deployment.​

Upgraded Deep Voice can mimic any voice in mere seconds

 

[cirr]

Robot maker plans expansion

2018-03-09 11:00 China Daily Editor: Li Yan

Visitors watch a medical robot designed by KUKA at the Appliance and Electronics World Expo in Shanghai on Thursday. (Photo/Xinhua)

Midea's robotic unit doubles production capacity to increase presence in China

Midea Group's robotic arm, in Germany, KUKA AG, will expand production capacity of robots and robotics systems in China, as the country's demand for the machines has been booming recently, according to a senior manager of the company.

"KUKA is working closely with Midea to look for business opportunities," said Wilfried Eberhardt, chief marketing officer of the company. "We will soon double the production capacity in China as the demand for robotics is increasing in the country."

The company's expansion plan was a move in accordance with a strategy of its parent company, which aims to promote the use of robotics in industrial and smart home appliances.

According to Eberhardt, the firm has started an expansion project in Shanghai, helping to double its capacity for robots and big systems.

In addition, the company will also increase its production capacity of a plant in Kunshan, Jiangsu province, for larger production of smaller systems.

Eberhardt said that KUKA's robots can be widely used in automotive, electronics, logistics, e-commerce, metal manufacturing and healthcare industries.

"Our robots are designed to make humans' life easier," he said.

On Thursday, KUKA displayed a number of medical robots at the Appliance and Electronics World Expo in Shanghai.

KUKA's KBR Med robot components provide precision, flexibility, safety, responsiveness and other features that can be effectively applied in orthopedic surgery, ultrasonic diagnosis and minimally invasive surgery by integrating the capabilities of the most acclaimed robots in the field of medical technology.

"Collaborative robots will be handy for home use by Chinese families," Eberhardt said.

According to Olaf Gehrels, general manager of Midea Robotics Company, the robotic industry will reach a scale of 100 million units by 2025, of which the Chinese manufacturers are predicted to take 50 percent.

"We are bringing together sensors, vision and voice, navigation, grasping, force control and more into one single user experience that bridges all our product areas－we call this the new era of the human-robot interface," Gehrels said.

"We believe that robotics and motion control are the future－robots will become (an) important part of logistics, medical and healthcare and even smart home appliances," he added.

http://www.ecns.cn/business/2018/03-09/295162.shtml

 

[TaiShang]

Midea Group's robotic arm, in Germany, KUKA AG, will expand production capacity of robots and robotics systems in China, as the country's demand for the machines has been booming recently, according to a senior manager of the company.

That's a good strategy to create jobs and more innovation in China.

Midea's acquiring of KUKA has been probably one of the most strategic moves in near acquisition history.

 

[onebyone]

A leading Chinese technology company, Baidu, has developed an artificial intelligence (AI) algorithm that can clone human speech immediately.

The system just needs 60 seconds of voice and can change a female voice to a male one. It can also turn a British accent into an American one. There is more; the AI can learn to mimic various styles of speaking and also personalize text-to-speech to a new level.

“From a technical perspective, this is an important breakthrough showing that a complicated generative modeling problem, namely speech synthesis, can be adapted to new cases by efficiently learning only from a few examples,” Leo Zou, a member of Baidu's communications team, told Digital Trends.

“Voice cloning is expected to have significant applications in the direction of personalization in human-machine interfaces,” the researchers mentioned in a Baidu blog article.

Baidu’s Deep Voice research team last year had unveiled it could clone voices with 30 minutes of training material. Interestingly, Adobe also has a similar program named VoCo that can mimic a voice with 20 minutes of audio.

Last year, Lyrebird used neural networks to mimic voices of global leaders including, President Donald Trump and former President Barack Obama.

Recently, a research report titled “Neural Voice Cloning with a Few Samples” was released by a team of researchers, who used two approaches for voice cloning: speaker adaptation and speaker encoding.

“We demonstrate that both approaches can achieve reasonable cloning quality even with only a few cloning audios,” researchers maintained.

Innovations in voice cloning in the recent years have also raised debate on its ethical basis. After large-scale use of Photoshop for fake news, voice cloning has left many worried.

Concerns are being raised over the possible use of voice cloning for blackmailing, financial frauds and similar illegal activities.

(With inputs from agencies)

https://news.cgtn.com/news/3367444e786b7a6333566d54/share_p.html

 

[war&peace]

This will have grave implications and legal issues for voice as an evidence in the court of law. DARPA already has been using this technology to produce the speeches of dead or non-existent terrorists.

 

[cirr]

AI-driven technology reshaping city traffic in China

2018-03-11 08:25 Xinhua Editor: Wang Fan

An ambulance in Hangzhou, an eastern Chinese city, avoids gridlock and all red lights switch to green as it approaches.

"The travel time was cut to half," said Sun Shixiang, of Hangzhou's public security bureau, while pointing to a big screen.

The progress is attributed to Alibaba's "City Brain", an AI platform on Alibaba's cloud infrastructure. It is a support to cities in digital transformation.

Around half of the world's population lives in urban areas. Traffic congestion goes along with urbanization.

In Hangzhou, the "City Brain" pulls in traffic and weather data and analyzes real-time traffic flow, adjusting traffic lights accordingly.

Over the past year, the technology has controlled traffic lights in 128 intersections in Hangzhou. The average speed of cars on these roads has increased by 15.3 percent, and travel time on bridges similarly reduced.

In addition to transport, the technology is expected to provide solutions for the city's energy and water supplies.

"City Brain" is active in cities such as Hangzhou, Suzhou and Quzhou. Outside China, Kuala Lumpur in Malaysia will soon adopt the system.

"Artificial intelligence will become important for all cities, helping them achieve sustainable development," said Wang Jian, Alibaba's technical committee chair.

 

[TaiShang]

China to publish guideline on AI development: minister
Xinhua, March 10, 2018

China will soon publish a guideline and detailed regulations on artificial intelligence (AI) development to make breakthroughs in critical technologies, Minister of Science and Technology Wan Gang said Saturday.


China is speeding up the research in the new generation of AI, Wan told a press conference on the sidelines of the annual session of the National People's Congress.


"We will promote AI application to address problems in security, health, environmental protection and so on," he said, adding the ministry will help domestic AI companies and research institutions find international cooperation.


China is also working on laws and regulations to tackle issues related to social ethics, job structure, personal privacy and national security, which may be brought by AI development, he said.


Last July, the State Council, China's Cabinet, issued a plan on new generation AI. The plan said the AI industry should be a major new growth engine and have improved people's lives by 2020, and set the target of China becoming a major center for AI innovation and leading the world in AI by 2030.

 

[Clutch]

Future Today Institute: China will become the world’s ‘unchallenged AI hegemon’ in 2018

• Khari Johnson

1 day ago

Future Today Institute founder Amy Webb speaks onstage at SXSW on March 11, 2018 in Austin, Texas

Future Today Institute founder Amy Webb has released her annual tech trends report, and much of it focuses on the continuing impact of artificial intelligence. Other trends highlighted by the report include space travel, human gene editing, and a global shortage of data scientists. Webb, a quantitative futurist and professor of strategic foresight at the NYU Stern School of Business, released the report today in a presentation at SXSW in Austin, Texas.

Now in its 11th year, the report identifies 225 trends across 20 industries, with roughly 70 of those trends related directly to AI.

In 2018, Webb expects the AI cloud and marketplaces for algorithms will continue to grow and the first personal robots will come to market.


China also garnered considerable attention in the report.

“The development of AI is our modern version of an arms race, and in 2018, China will lay the groundwork to become the world’s unchallenged AI hegemon. If data is the new oil, China’s massive, 730 million online population puts it in control of our largest, and possibly most important natural resource going forward — human data — and it doesn’t have the privacy and security restrictions that might hinder progress in other nations,” the report says.

China has invested more sovereign state funding than all other nations combined and is the single biggest investor in AI startup companies, Webb said.

“I don’t begrudge China as a nation’s ability to do this, but I do question what their end-game is and whether that’s good for the future of humanity,” she said.

Facial recognition, voice recognition, and personality recognition — all influenced by AI — were also listed among trends expected to continue.

Recognition tech, AI assistants like Siri, and trends like dedicated AI chips mean 2018 will be the beginning of the end for the traditional smartphone, Webb said, and signal increased augmentation through wearables like smart watches, smart glasses, and smart earbuds.

Digital assistants like Siri and Google Assistant are becoming ubiquitous, and Webb predicts that by 2021 more than half of all computing will be done with voice.

After tabulation and programmable systems, AI defines the third era of computing, the report finds. Consolidation and acquisition of startups by tech giants will also continue in the year ahead.

“These are the big nine companies that control the future of AI,” Webb said. “They’re all partnering with academic institutions, the money’s flowing from here, there’s a free flowing of research going back and forth, and so essentially our next era of computing is in the hands of these nine companies and three of those companies — Tencent, Baidu and Alibaba — I would argue are probably the most important ones you should pay attention to.”

 

[cirr]

China boosts home appliance industry upgrading

2018-03-13 10:14 CGTN Editor: Gu Liping

China's home appliance industry is gradually becoming a global leader in the field of AI and Internet of Things (IoT), according to a report released at the Appliance & Electronics World Expo (AWE) in Shanghai.

More than 250,000 visitors and over 800 home appliance and electronics manufacturers participated in the AWE, which ended on Sunday.

All leading Chinese home appliance makers joined in promoting intellectualization in upgrading traditional manufacturing operations.

Haier Group, China's largest home appliance maker, has set up the country's first national-class industrial Internet platform COSMOPlat and the world's leading Internet platform that has independent intellectual property rights in China.

In 2017, COSMOPlat became the world's largest mass customized solution platform with total transactions reaching 313.3 billion yuan (49.42 billion US dollars) and 4.12 million orders.

The world's biggest air-conditioner maker Gree Electric Appliances has been focusing on converting their existing eight production bases into unmanned factories in a push to upgrade manufacturing techniques with technology.

Local brand Aucma, previously known as a leading maker of refrigerators and freezers, began to develop advanced equipment in recent years, such as ultra-low temperature cooling systems and driverless trucks equipped with a dual vision system and AI driving system.

Midea Group's robotic arm, in Germany, KUKA AG, will double the production capacity of robots and robotics systems in China, as the country's demand for robotics has been booming recently. Midea Group has announced M-Smart, a smart home system designed for the future of family living.

In the past few years, the company has invested 20 billion yuan into research and development of smart home solutions, including the establishment of 17 research centers in eight countries, with more than 10,000 employees involved in R&D and 26,000 authorized patents.

"Midea Group sold over 300 million products each year; that's a huge base for our AI 'smart home' platform," said Zhang Xiaoyi, CIO of Midea Group. "Therefore we are developing products and services that go beyond the household into the community and 'new retail' scenario, to realize our goal of 'connecting people,' instead of merely connecting home appliances."

According to statistics from market consultancy Statista, China's smart home market is expected to reach a value of 130 billion yuan (20.3 billion US dollars) by the end of 2018, with an annual growth rate of about 48 percent, a big jump from the 40.3 billion yuan market in 2015.

 

[cirr]

Hikvision's AI roadshow attracts visitors in Bucharest

2018-03-15 09:52 Xinhua Editor: Gu Liping

An AI roadshow offered on Wednesday by Hikvision, a world leading provider of video surveillance products and solutions based in Hangzhou, China, attracted local city managers, as well as ordinary visitors in Bucharest.

Visitors climbed aboard Hikvision's mobile showroom, a long truck parked aside the Red Dragon Mall, northeastern Bucharest, to experience and discover technologies, such as face recognition, and see live demonstrations as well as presentations on cutting-edge topics such as cyber security.

"Hikvision offers almost all range of surveillance products and all these products can be delivered as an integrated security solution package, so as to help system integrators to save time and money while implementing projects," a local employee of the company told visitors.

"During our roadshow, we want to show people our latest products and customized solutions, with the aim to reflect Hikvision's global leadership in the video surveillance market through innovative products, pioneering solutions and prompt service," Jermy Wang, Romania Country Manager told Xinhua, adding that the company is committed to providing the latest technologies with the best quality products.

A local AI engineer said that the awareness about security products and solutions is growing in Romania, and he is optimistic about the prospect of Hikvision products in the local market.

"In terms of traffic intelligence management, Romania is still in its infancy. In major cities, including the capital, few cameras are installed in the main traffic lanes and intersections, showing the great potential of the market," Jin Xiaozhong, chairman of Yiwu Association in Romania, who is actively promoting several cities to adopt the smart traffic and smart city solutions of Hikvision.

Three separate trucks of Hikvision are travelling now through various European regions, visiting over 71 cities, said the company.