Sensory&Philips-Enhance ASR with Speech Enhancement

Sensory, a Silicon Valley company enhancing user experience and security for consumer electronics, announced today its collaboration with Philips, a provider of advanced speech enhancement technologies, to offer a combined technology suite. This would package Sensory’s best-in-class speech recognition technologies TrulyHandsfree™ and TrulyNatural™ with Philips BeClear Speech Enhancement™ algorithms, resulting in significant accuracy improvement in noisy environments. By processing an audio signal with Philips’ echo cancellation, noise suppression and/or beam-forming processors before passing it to Sensory’s speech recognition engine, much of the unwanted ambient noise in a signal can be filtered out, leaving the critical speech portion of the signal largely untouched. This process allows Sensory’s already noise robust speech recognizer to decipher near- and far-field speech more accurately in conditions where very high ambient noise is present.

“We tested various implementations of our BeClear Speech Enhancement algorithms with Sensory’s industry-leading speech recognition engine and have seen impressive improvements in accuracy in noiseless to high-noise environments,” said Navin Natoewal, General Manager, Media Interaction at Philips. “Manufacturers are focused on improving the user experience of their products, and many plan to accomplish this by implementing a hands-free interface that is effective in noisy environments. We are excited to be working with Sensory to offer a packaged solution featuring the industry’s most accurate embedded speech recognition technologies with our effective BeClear Speech Enhancement suite, a combination that will ensure that devices hear their users even at far distances in the noisiest environments.”

Philips’ BeClear Speech Enhancement algorithms have been tested and proven to provide increased efficacy for Sensory’s TrulyHandsfree and TrulyNatural speech recognition technologies. This combined offering of Philips and Sensory technologies has applications in multiple categories of consumer electronics, including TV and home theater, mobile devices, digital assistants, Internet of Things, smart homes, automotive, as well as others.

“Without speech enhancement added to the equation, Sensory proudly provides the most noise-robust speech recognition technology available to the industry,” said Bill Teasley, Vice President of Engineering of Sensory. “However, for many implementations like mobile devices, automotive and even smart homes where ambient noise can drown out speech, adding speech enhancement algorithms from an industry leader like Philips to the equation can greatly improve the efficacy and accuracy of our speech recognition in noise. This combination offers a significant benefit to the end user.”

About Philips Beclear-

https://www.philips.com/a-w/about/innovation/ips/ip-licensing/programs/beclear-speech-enhancement.html

BeClear Speech Enhancement

• One flexible suite of algorithms used for simultaneous voice communication and speech recognition for voice control usage
• Optimized for commercially available MCU and DSP platforms
• Manufacturer tuning to specific hardware form factor
• Tunable to any speech recognition (SR) engines without insight into SR pre-processing
• Optimized per use case (smart phone, smart watch, TV and voice control for IoT)
• Speeds up integration for new use cases
• All use case extensions based on proven technology

Includes the following Audio Front End processing technology suite -

• Super de-reverberationn
• Multi-channel echo cancellation
• Fast tracking beamforming
• Dynamic noise supression
• Far-end signal enhancement