June has been a busy month here at Webex. We have just announced a range of new features to make hybrid work more inclusive, supportive, and flexible. In the Collaboration AI team I lead as VP of Engineering, we support efforts with market-leading AI applications that allow our users to work from anywhere on any device. We have introduced applications in Webex Assistant, video background replacement, gesture recognition, improved transcriptions, and speech enhancement.
We continue to work on key AI applications to empower the future of work, including exciting innovations in speech AI.
Webex audio intelligence — the next evolution in speech communications
Since the emergence of digital network-based voice communications, the essential requirements were pretty simple – capture speech in real-time with microphones, encode it for transmission, deliver it to the endpoints, decode it and play it back for the listeners. All the building-block functions — power normalization, echo removal, coding to save bandwidth and resist transmission packet losses, and adaptation to myriad output devices — have evolved steadily but without much excitement over the past quarter-century. But the real excitement is now happening as we bring true audio intelligence to communications and collaboration.
Speech streams have a wealth of information about intent, identity, location, accent, and other details intertwined in the audio streams. The availability of cheaper, real-time computing capacity, the emergence of neural-network-based algorithm design, and the rethinking of human-machine interface design changes everything. Cisco is making broad, sustained investments in audio intelligence to power the next generation of Webex. Given how natural and efficient it is to use voice commands (a.k.a, natural language) to interface with machines to manage our day-to-day life, the use cases for speech AI have grown significantly.
Speech AI – improving collaboration and meeting administration
Below are some of our recent innovations improving live collaboration experiences, as well as pre-meeting and post-meeting experiences:
- Webex Assistant for Devices — More languages, skills, and toolkits for developers
Webex Assistant for Devices offers zero-touch, voice-activated meeting and device controls to serve the workplace’s most common meeting room needs. We are now bringing this capability to our partners and app developers to expand user experiences and support more use cases by enabling the Webex Assistant Skills platform. This includes documentation, APIs, toolkits to build custom applications for their own organizations and their users or customers by tapping into the intelligence of Webex Assistant. Also, now all our users can configure Webex Assistant for devices to engage in English, French or Spanish.
- Webex audio intelligence — “optimize for my voice” speech enhancement feature
We live in a noisy world, but that shouldn’t prevent us from having engaging meetings. In October 2020, we announced our superior speech enhancement and background noise removal technology with the acquisition of BabbleLabs. This month we crossed a milestone by removing background noise from 16 billion meeting minutes. With the Webex audio intelligence: optimize for my voice feature, we go beyond removing background noise with innovations focusing on de-reverberation, enhancing speech quality, and speech separation. Webex now distinguishes between foreground and background talkers. When you select “optimize for my voice” in your meeting, only voices closest to the microphone are captured, and all background speech is removed — while simultaneously removing all background noise and room reverberation that may otherwise affect clear speech.
We will enable ever more precise handling of each talker — studio sound in any environment, personalization to produce and consume audio that best reflects who you are or want to be, analytics for real-time immersion in people, and personal insights.
Other AI innovations
Earlier this year, we announced our people insights feature. This valuable functionality provides real-time data (private for each individual) to enable smarter, more informed decisions about our meetings. We enhanced this capability with the introduction of personal insights (also private for each individual) that enables setting personal goals within the Webex Suite. Webex customers can find more information in Cisco Customer Connection.
In addition, we continue to make significant improvements with our virtual background replacement feature to improve the experience and enable it on all platforms and devices. And we’re deploying fast, accurate, touchless gesture recognition to make meeting responses even more intuitive.
To catch up on all we have in our portfolio today and read about our team members and our upcoming innovations, visit the AI-Enabled Experiences for Webex page.
Learn more
16 billion minutes of noisy distractions disappeared!
What’s new in Webex: June 2021
Life happens. Background noise doesn’t have to
The rise of AI speech enhancement & better team collaboration
About The Author
