Baidu’s voice cloning AI, Alexa’s creepy glitch, and exciting health tech released at HIMSS. This and more in our weekly news roundup
Our carefully curated selection of articles from this week on CliqueApi and other respected news sources from around the world.
Otter, the new app which can transcribe voice conversations in real-time, Clique
Otter, a new voice technology app designed by former Google architect Sam Liang and released at this year’s Mobile World Congress, provides an easier way for users to transcribe voice conversations and browse through them whenever they wish. The app was designed to understand and transcribe conversations between multiple human speakers, allowing users to keep a written record of meetings, interviews, conferences, and every day conversations.
In with the bricks. The smarthome pioneer adding voicetech as standard, Clique
Quantify Technology, an Australian company specializing in smart buildings tech, is currently integrating Amazon Alexa devices into new homes and apartments, allowing tenants to control their lights, heating, cameras, locks, media, and other household utilities with their voice. Rather than installing Amazon’s Echo speakers, Quantify will be using its very own Alexa-powered smart speakers, which come with the company’s Alexa skill for voice-controlled homes.
HIMSS 2018: Healthcare sector companies showcase exciting new tech, Clique
This year’s Healthcare Information and Management Systems Society (HIMSS) conference took place from March 5 to March 9 at the Venetian Palazzo, Sands Expo Center, in Las Vegas. The event was a chance for companies and developers to present a variety of new interesting tech for healthcare settings, including collaboration tools, voice-assistants, and deep learning algorithms. Clique reports on some of the exciting new platforms and services released at at HIMSS18.
For real. Baidu makes enormous leap forward in AI generated voicetech, Clique
Researchers working at Chinese tech giant, Baidu, have developed new voicetech that can reproduce anyone’s voice after listening to it for just 3.7 seconds. Around the same time last year, the company’s same voice cloning system, Deep Voice, needed to process around 30 seconds of audio before it could effectively create a duplicate of the speaker’s voice. While highly-performing neural voice cloning can have many useful applications, it could also have serious ethical and security implications.
Omega World Travel launches Alexa skill for travel managers, PhocusWire
Travel company, Omega World Travel, has recently developed an Alexa skill for Echo Devices and is soon planning to roll out more voice-enabled services for agents and consumers alike. The company’s first-ever Alexa skill provides duty-of-care information for travel managers and for Omega’s own operations team, drawing from real-time data that the company had previously made accessible to its clients in the form of searchable dashboards.
Visually impaired students learn to code with talking technology, Kxan
On Wednesday, Apple representatives visited Texas School for the Blind, unveiling a new app designed to assist visually impaired students in building their own computer programs. Swift Playground reads descriptions to the students out loud, allowing them to make preferred selections via voice and effectively program drones. A group of students tested the app during school hours, identifying a few glitches in the system that leave space for future improvement.
Amazon says the laughing Echo devices that terrified some users happened because Alexa ‘mistakenly’ thought it heard instructions to laugh, Business Insider
Some Amazon Echo users had reported sudden and creepy fits of laughter by voice assistant Alexa. Amazon confirmed this on Wednesday, adding that these unusual laughing incidents were caused by the devices mistakenly hearing the command phrase “Alexa, laugh”. The company is now changing the laugh-triggering phrase to ‘Alexa, can you laugh?’, which is said to be less likely to have false positives and prompt the unasked-for unnerving chuckling episodes.
ELSA raises $3.2M for it’s A.I.-powered English pronunciation assistant, TechCrunch
ELSA, which stands for English Language Speech Assistant, is a new language learning app that recently raised $3.2 million in funding. The AI-powered platform uses speech recognition technology to help new language learners to improve their pronunciation. The San Francisco-based startup that developed the app was founded in 2015 by Stanford graduate Vu Van and Dr. Xavier Anguera.
Singapore is a global leader in fingerprint biometrics and voice recognition technologies in the financial services sector: Survey, Business Insider
A recent International study commissioned by identity data intelligence firm, GBG and carried out by advisory company Forrester Consulting, found that Singapore is now a global leader in fingerprint biometrics and voice recognition technologies. The study surveyed 315 companies in several countries, including Singapore, China, the UK, Australia, and the US. The collected results revealed that Singapore is ahead of all other countries in the adoption of both fingerprint biometrics and voice recognition tools.
Cork’s Velona Systems Could make VoIP a lot more secure, Irish Tech News
At this year’s Mobile World Congress in Barcelona, an Irish startup specialized in cyber-security solutions released a new tool that could improve the security of VoIP services. The company’s latest technology is one of the first security measures to focus on session initiated protocols, or SIP, allowing users to detect potential attacks on their VoIP systems. Velona Systems’ new technology could, among other things, prevent malicious users from using other people’s accounts to make long-distance phone calls.