NEC develops automatic optimization technology for deep learning

In an effort to facilitate improvements in recognition accuracy, NEC has developed automatic optimization technology for deep learning. In a statement, the company explains that if deep learning systems become excessively familiar with data, they become unable to accurately recognize data that they have not learned. This “overtraining” results in degradation of recognition accuracy when dealing with data that was not used in the learning process. To prevent overtraining, “regularization” technology is commonly used, which regulates the extent of learning

Alibaba speech and facial recognition tech to be deployed in Shanghai subway system

Speech and facial recognition technologies developed by Alibaba will be deployed in Shanghai subway system, according to a report in the South China Morning Post. Alibaba and Shanghai Shentong Metro Group said this week that the Metro would install “far-field” speech recognition technology in ticket machines in all stations, as well as facial recognition systems at station entrances to verify commuters identities. According to Yan Zhijie, director of intelligent speech recognition at Alibaba’s Institute of Data and Science Technology, far-field

Google and Stanford researchers developing voice recognition models for medical conversations

Google is working with researchers and physicians at Stanford University on Automatic Speech Recognition (ASR) models to transcribe medical conversations. In a blog post, Katherine Chou, Product Manager and Chung-Cheng Chiu, Software Engineer, Google Brain Team say that Electronic Health Records (EHR) documentation often takes half of a doctor’s 11-hour workday, contributing to burnout. Google cites recent research published in the Annals of Family Medicine which shows that physician satisfaction and medical chart quality and accuracy improved when scribes took

UK firm XMOS raises $15M in series E funding round

UK voice capture solutions firm XMOS Ltd recently raised $15M in a Series E funding round led by Infineon Technologies. In addition, existing investors Amadeus Capital Partners, Draper Esprit, Foundation Capital and Robert Bosch Venture Capital participated in the funding. The company said it intends to use the funds for its product development plans. Led by president and CEO Mark Lippett, XMOS provides a range of voice and audio solutions to the consumer electronics market by combining voice processing, biometrics and

Conexant partners with Baidu on voice-based AI devices

Conexant, a leading provider of audio and voice technology solutions that enable a more natural user experience, will work with Baidu to release development kits and reference designs for device makers to develop voice-enabled artificial intelligent (AI) devices running on Baidu’s DuerOS platform. DuerOS is a conversation-based AI system that supports access to a voice-activated digital assistant for mobile phones, TVs, speakers, and other devices. “The voice revolution is a global phenomenon. By working with Baidu, we help more third-party

Sensory integrates AI into TrulyNatural speech recognition solutions

Sensory has integrated artificial intelligence (AI) technology into its TrulyNatural line of embedded natural language speech recognition engines. TrulyNatural is now embedded with a range of neural net deep learning methods such as wake word technology for Alexa and other cloud engines, natural language voice interfaces, biometrics, and computer vision. The company’s AI technology lineup includes TrulyHandsfree, wake words and small vocabulary low power solutions; TrulySecure, biometric authentication with face and/or voice; and TrulyNatural, an embedded natural language large vocabulary

Apple developing a processor to improve the way devices handle AI tasks

Apple Inc. is developing a processor that will improve the way the company’s devices handle artificial intelligence (AI)-related tasks, such as facial recognition and speech recognition, according to a report by Bloomberg that cited an individual familiar with the matter. Apple Neural Engine is designed to make significant improvements to Apple’s hardware over time, and will eventually be integrated into the iPhone, iPad and other devices, according to the source. The company has already tested the chip with prototypes of

Sensory releases v5 of TrulyHandsfree speech recognition system

Sensory has released version 5.0 of its TrulyHandsfree speech recognition system, which includes significant updates to the embedded AI to significantly increase its performance and accuracy while staying small and low power. Incorporating Sensory’s smartest and most efficient deep neural network technologies to date, TrulyHandsfree 5.0 delivers an on-device voice user interface experience that is more natural and intuitive than previous versions. The product’s new shallow learning approach reduces the model sizes to operate in ultra-low power and with minimal

IBM researchers achieve new records in speech recognition

IBM researchers have set a milestone in conversational speech recognition by achieving a new industry record of a 5.5 percent word error rate, surpassing its previous record of 6.9 percent, according to the company’s blog post. The researchers conducted a difficult speech recognition task to achieve this record, where they recorded conversations between humans discussing typical everyday topics like “buying a car.” This recorded corpus, titled “SWITCHBOARD”, has been used for over two decades to benchmark speech recognition systems. To

Wells Fargo working on voice-based payment capabilities

Wells Fargo & Co. is developing a voice biometrics-based payments offering for consumers that will reportedly expand beyond the basic day-to-day functionality offered by current virtual financial assistants, according to a report by DigiDay. “These digital assistants like Siri or Alexa — these things are just starting,” Steve Ellis, head of Wells Fargo’s innovation group, said. “There’s a really big future here for how our customers interact with us. We are starting to do proofs of concept with information exchange