Lots of news out of China this week. Biggest story was Victor Gevers, a security researcher for the GDI Foundation, finding a database belonging to SenseNets. The Shenzen based company provides facial recognition and other monitoring systems to the Chinese police.
Everyone knows that China is utilizing AI to enhance the government’s surveillance capabilities, but this was the first insight we’ve gained into the scope of the government’s operations in the Xianjiang province. It’s looking quite Orwellian there. It’s also notable that SenseTime divested their stake in SenseNets right after finding out about the security breach.
Company developments:
Levi’s Hires Chief Artificial Intelligence Officer – Feb. 22, 2019 (Total Retail)
- Levi’s has hired its first artificial intelligence (AI) officer. Katia Walsh, who was most recently Vodafone’s chief data and analytics officer, will begin her tenure at Levi’s on April 29
- According to a company statement, Walsh will focus on building “data, analytics and artificial intelligence enablers” that will strengthen the company’s present and future business ventures. Levi’s hasn’t been more specific than that, but it’s a safe bet Walsh will tap her expertise to build on the AI efforts specifically — and the digital advances in general — Levi’s has made in recent years
Waymo driverless cars now understand hand signals from police officers – Feb. 22, 2019 (Gear Brain)
- The new skill is shown off in a video published to Waymo’s YouTube account this week
- As Waymo explains in the video description: “The video shows Waymo’s self-driving car approaching a traffic light that is not working. The car comes to a stop before entering the intersection and waits for the officer signal that we can proceed.”
Google AI technique reduces speech recognition errors by 29% – Feb. 21, 2019 (VentureBeat)
- Speech recognition is pretty darn good these days. State-of-the-art models like EdgeSpeechNet, which was detailed in a research paper late last year, are capable of achieving about 97 percent accuracy. But even the best systems sometimes stumble on uncommon and rare words
- To narrow the gap, scientists at Google and the University of California propose an approach that taps a spelling correction model trained on text-only data. In a paper published on the preprint server Arxiv.org (“A Spelling Correction Model for End-to-End Speech Recognition“), they report that in experiments with the 800-word, 960-hour language modeling LibriSpeech dataset, their technique shows a 18.6 percent relative improvement in word error rate (WER) over the baseline. In some cases, it even managed 29 percent error reduction
Facebook is Working on Its Own Custom AI Silicon – Feb. 20, 2019 (Extreme Tech)
- Facebook’s chief AI researcher, Yann LeCun, has stated that the company is working on its own custom AI silicon, with the goal of building far more efficient methods of processing neural networks in hardware and boosting performance, addressable problems, and energy efficiency
- “We don’t want to leave any stone unturned, particularly if no one else is turning them over,” he said in an interview before presenting a paper on the history and future of machine learning at ISSCC (International Solid State Circuits Conference) in San Francisco. Exact details on what Facebook is building remain vague, though Intel announced an AI-focused partnership with the company at CES this year
Fundraising / investment:
Nreal raises $16 million for lightweight mixed reality glasses – Feb. 22, 2019 (VentureBeat)
- The money came from China Everbright Limited New Economy Fund, iQiyi, Shunwei Capital, and China Growth Capital. The latest round comes just one month after the company announced a $15 million round at the CES 2019 tech trade show in Las Vegas. That was where I saw a demo of the company’s high-quality mixed reality glasses
- “China Everbright Limited New Economy Fund has always been keen on the investment on those start-ups who are tech-driven and capable to commercialize its products,” said Victor Ai, managing director of China Everbright, in a statement. “In less than three years, we have been growing into an important force in private equity market, specializing in ‘AI+’ investment
Splice Machine raises $16 million for unified machine learning platform – Feb. 20, 2019 (VentureBeat)
- Splice Machine, a self-described data platform for “operational” artificial intelligence (AI), today announced that it’s raised $16 million in series B financing led by GreatPoint Ventures (GPV), with minority equity participation from Accenture Ventures and individual investor and SunBridge Group CEO Allen Miner
- The influx of fresh capital — which follows on the heels of a $9 million round led by Salesforce Ventures in December 2017, and which brings Splice Machine’s total raised to $58 million — will be used to “further innovate” its product portfolio and grow its engineering and sales teams, according to CEO Monte Zweben
Armorblox raises $16.5 million for NLP that secures human communication – Feb. 20, 2019 (VentureBeat)
- Natural language processing (NLP) — the branch of AI that helps computers understand and manipulate human language — might not seem naturally applicable to cybersecurity, but as it turns out, the benefits can be enormous. Just ask Dhananjay Sampath. He’s the cofounder of Armorblox, a cybersecurity startup that emerged from stealth today with $16.5 million in venture capital backing led by General Catalyst, along with Microsoft chairman John Thompson, former U.S. chief data scientist DJ Patil, Google vice president of security engineering Gerhard Eschelbeck, and other individual investors
- The San Francisco startup’s technology, which it plans to showcase at the upcoming RSA Conference in March, taps NLP to analyze sensitive information in emails and documents and to protect against data- and identity-related attacks
ViSenze raises $20 million to further develop its AI product comparison tools – Feb. 19, 2019 (VentureBeat)
- The Singapore startup, which offers an image recognition platform that enables store customers to search for products by image instead of keyword, counts Urban Outfitters, Uniqlo, Zalora, Rakuten, and “major” original equipment manufacturers (OEMs) among its stable of customers, and says it’s linked over 400 million products from 800 merchants and brands on its global affiliate network
- After three years of steady growth (200 percent) that saw the number of users conducting searches with its products exceed 300 million (and search volume surpass three million queries a day), ViSenze is raising venture capital to further “develop [its] capabilities” and network and “[expand its] global reach,” according to CEO and cofounder Oliver Tan
Senseon raises $6.4M to tackle cybersecurity threats with an AI ‘triangulation’ approach – Feb. 19, 2019 (TechCrunch)
- Senseon, which has pioneered a new model that it calls “AI triangulation” — simultaneously applying artificial intelligence algorithms to oversee, monitor and defend an organization’s network appliances, endpoints and “investigator bots” covering multiple microservices — has raised $6.4 million in seed funding
- David Atkinson — the startup’s CEO and founder who had previously been the commercial director for Darktrace and before that helped pioneer new cybersecurity techniques as an operative at the U.K.’s Ministry of Defense — said that Senseon will use the funding to continue to expand its business both in Europe and the U.S
AI, big data company AiSensum receives seed funding from 500 Startups – Feb. 18, 2019 (Yahoo)
- AiSensum, an AI and big data analytics company with presence in Singapore, Jakarta, and Bangalore, announced today that it has received seed funding from 500 Startups. The amount raised is unrevealed
- Launched in February 2018, AiSensum designed an analytics platform called OctoPi, which essentially is an Artificial Intelligence/Machine Learning driven platform that creates additional revenue streams for enterprises and startups through Data Monetisation Partnerships
Facebook chief AI scientist joins $4.5 million round for algorithmic music startup Antescofo – Feb. 18, 2019 (VentureBeat)
- The artificial intelligence company is known for its classical music lesson mobile app Metronaut. Other investors in the round include venture firms Daphni and OneRagTime, plus angel investors Nobuyuki Idei, former Sony Corporation CEO and chairman; Sophie Gasperment, a senior executive at L’Oréal; and Thibault Viort, CEO of new businesses for AccorHotels Group
- “The new funds will allow us to strengthen our position as an innovative actor in the music industry, grow internationally, notably into China, and will enable Antescofo to build up Metronaut’s catalog of accompaniments, so we can satisfy our users,” said Antescofo CEO Laurent Tran Van Lieu in a statement
Partnerships:
Clobotics joins forces with GEV to bring AI to the wind sector – Feb. 22, 2019 (Renewable Energy Magazine)
- The global partnership will use Clobotics pioneering artificial intelligence solutions to deliver the blade inspections, fuelling the ongoing digitalisation of the global wind O&M market. The partnership will cover all the regions GEV currently serves, bringing Clobotics’ solutions to a broad swath of the globe, including Europe, Africa and North America
- With Clobotics Smart Wind, the autonomous drone flies around the turbine and builds a 3D picture of all the blades, capturing high resolution images as it flies, and then lands back to the starting point upon completion. Inspection is done in minutes, rather than hours. Images of blades are then automatically analyzed, annotated, and reported on the cloud-based customer portal the very next day. Using the portal, customers can easily track the entire lifecycle of turbines, see the progression of damage and view trend reports on individual turbines or their entire fleet
CEA-Leti & Stanford Target Edge-AI Apps with Breakthrough NVM Memory Cell – Feb. 20, 2019 (Design & Reuse)
- Researchers at CEA-Leti and Stanford University have developed the world’s first circuit integrating multiple-bit non-volatile memory (NVM) technology called Resistive RAM (RRAM) with silicon computing units, as well as new memory resiliency features that provide 2.3-times the capacity of existing RRAM
- The proof-of-concept chip has been validated for a wide variety of applications (machine learning, control, security). Designed by a Stanford team led by Professors Subhasish Mitra and H.-S. Philip Wong and realized in CEA-Leti’s cleanroom in Grenoble, France, the chip monolithically integrates two heterogeneous technologies: 18 kilobytes (KB) of on-chip RRAM on top of commercial 130nm silicon CMOS with a 16-bit general-purpose microcontroller core with 8KB of SRAM
Accenture Helps Metro de Madrid Balance Energy Efficiency and Passenger Comfort with AI-Based Self-Learning Ventilation System – Feb. 19, 2019 (AP News)
- Accenture (NYSE: ACN) has helped Metro de Madrid develop and implement a self-learning AI-based ventilation system that minimizes energy costs and emissions and ensures high air quality in metro stations and commuters’ comfort
- The artificial intelligence (AI)-based system has enabled Metro de Madrid to reduce its energy costs for ventilation by 25 percent and cut CO 2 emissions by 1,800 tons annually. Accenture will present the system at this year’s MWC (formerly known as Mobile World Congress) in Barcelona, Feb. 25-28
Research / studies:
Scientists Developed an AI So Advanced They Say It’s Too Dangerous to Release – Feb. 19, 2019 (Science Alert)
- A group of computer scientists once backed by Elon Musk has caused some alarm by developing an advanced artificial intelligence (AI) they say is too dangerous to release to the public
- OpenAI, a research non-profit based in San Francisco, says its “chameleon-like” language prediction system, called GPT–2, will only ever see a limited release in a scaled-down version, due to “concerns about malicious applications of the technology”
AI detects potentially damaging ice on wind turbines – Feb. 18, 2019 (VentureBeat)
- Ice is the enemy of turbines everywhere. Some wind farms report energy production losses of up to 20 percent due to icing, according to Canadian wind-industry consultancy firm TechnoCentre Éolien (TCE), and that’s not the worst of it. Over time, ice shedding from blades can damage other blades or overstress internal components, necessitating costly repairs
- There’s a clear and present use case, then, for an AI system that detects wind turbine icing. Fortunately, that’s just what a team of researchers recently described in a paper published on the preprint server Arxiv.org (“WaveletFCNN: A Deep Time Series Classification Model for Wind Turbine Blade Icing Detection“).
Government / policy:
China has turned Xinjiang into a zone of repression — and a frightening window into the future – Feb. 23, 2019 (Washington Post)
- AT A minimum, the minority Muslim Uighur population of Xinjiang province in China is about 11 million people, and probably significantly higher. So consider the scope of surveillance over Uighurs in light of a recent database leak that indicated about 2.5 million people in Xinjiang are being tracked by cameras and other devices, generating more than 6.6 million GPS coordinates in one 24-hour period, much of it tagged with locations such as “mosque” and “hotel.”
- Victor Gevers, a security researcher for the GDI Foundation, a nonprofit that seeks to defend Internet freedom, found the database, belonging to SenseNets, a Chinese company that provides facial recognition and other monitoring systems to the police
UK AI investment tops 1bn in 2018 as the number of start-ups grows – Feb. 21, 2019 (Compelo)
- The new figures from Tech Nation and Digital Economy Council mean that British companies raised more capital than France, Germany and the rest of Europe combined
- Graphcore was one of the biggest recipients of funding last year, raising $200m (£153m) from a consortium of investors. The Bristol-based start-up is currently valued at $1.7bn (£1.32bn) and is looking at ways to improve the speed of AI processing and develop technologies for the next generation of AI enterprises and innovators
Chinese state media’s latest innovation is an AI female news anchor – Feb. 20, 2019 (Quartz)
- State news outlet Xinhua announced yesterday (Feb. 19, link in Chinese) that it had, in collaboration with search engine Sogou, created the world’s first female AI news anchor, known as Xin Xiaomeng. The anchor will make “her” debut during the upcoming Two Sessions political meetings at the start of March
- Xinhua and Sogou said that they also developed an improved male anchor called Xin Xiaohao, who is also able to stand up and gesticulate and has more natural mouth movements
Chengdu offers AI startups RMB 3 million subsidies as race for talent heats up – Feb. 18, 2019 (TechNode)
- The southwestern Chinese city of Chengdu is looking to drive innovation in artificial intelligence (AI) by offering subsidies of up to RMB 3 million (around $430,000) to startups in the city
- Chengdu hopes to establish itself as an AI-powered innovation center, covering a variety of applications including 5G, ultra high-definition video streaming, and virtual reality. The city government is offering subsidies of RMB 3 million to companies that innovate in a number of AI fields, including medical and military technologies, according to a statement released on Feb. 15 (in Chinese)