Whether it's removing the background noise of your dog's barking or blurring out a home office renovation in progress, AI-based cloud video features are as useful as they are innovative. But what if AI could take a more proactive role, making suggestions before you even think about scheduling a meeting, or providing important relevant content without prompting? These may seem like science fiction wishful thinking, but as AI advances, anything is possible.
AI is a key factor driving the cloud video services market forward, and all major providers have rich AI-based feature sets, whether developed in-house, acquired, or integrated. Meanwhile, some cloud video providers are turning to alternative sources of AI video capabilities, namely cloud-native suites of AI capabilities accelerated with graphical processor units provided by GPU vendor Nvidia. Nvidia's service, called Maxine, allows providers to bring features like facial alignment, bandwidth efficiency, and custom avatars to their video services, said UC analyst Zeus Kerravala. No jitter article.
Regardless of how you incorporate AI, as Nema Prachi, senior analyst for enterprise communications, writes in her recent book, much innovation is still to come. Omdia Universe Report “Collaborative Meeting Services, 2021”. Nema said he expects CIOs to increasingly seek AI-based capabilities to meet their cost-cutting and productivity-enhancing needs, and to see “tremendous innovation in simplifying the user experience before, during and after meetings.”
It is in this future of innovation that AI will have real value, two industry analysts agreed.
What is AI? More than just automation
said Ira M. Weinstein, Founder and Managing Partner of Recon Research. The pandemic has also accelerated the use of video conferencing, so cloud video is now almost synonymous with enterprise communication and collaboration.
Weinstein said cloud video providers have significantly increased capacity to meet increased demand during the pandemic. Recent advances in processing power, machine learning, and big data management have aligned the stars to allow providers to expand AI-based capabilities within their platforms, Weinstein added. Cloud video vendors now have access to more experienced AI developers than ever before, which also supports their AI efforts, Weinstein noted.
And of course, it's all about data, says Dave Michels, Principal Analyst at TalkingPointz. “AI thrives on data. [always] “We need more data for context and…learning”—the more data an AI engine has, the easier it is to find correlations and the farther it can predict events. For example, Michels detailed that advanced AI capabilities allow users to tell a voice-activated in-room video system to “call Ryan,” and the AI can determine which of the hundreds of Ryans in the company directory makes the most sense based on who made the request and previous interactions.
As Weinstein said, what makes AI different is its ability to make decisions. “AI doesn't just automate functions that humans do, it analyzes information, situations, and context and makes decisions based on that,” Weinstein said.
Introducing AI: Find adoption for transcription, translation
While we may be far from the future of AI in meetings, cloud video vendors are adding a number of AI capabilities to their platforms. As noted in the Omdia report, these include translation, closed captioning, virtual meeting assistance, transcription, voice control, meeting summaries and follow-up items, the ability to retrieve action items, meeting scheduling automation, people insights from social media, discussion and participant tagging.
Another common AI feature that adds value to corporate meeting experiences is noise reduction and cancellation, which uses algorithms to improve audio quality or reduce background noise, according to Weinstein. He added that Cisco, Google, Microsoft, Zoom and others have implemented some form of noise cancellation into their cloud video platforms. for example, zoom In 2018, we developed noise canceling for Zoom meetings. In the past year, Cisco has acquired technology from the following companies: bubble lab For use in Webex meetings. and added by microsoft Last year, we introduced noise suppression to Teams.
Weinstein said audio transcription is also important. Similarly, Cisco built transcription capabilities into Webex through acquisition. voicea In 2019, Microsoft rolled out meeting transcription for Teams last year, and Zoom announced its Zoom meeting transcription feature in 2019. zoomtopia 2017. In addition, users can utilize transcription services from third-party vendors such as: Otter.ai.
Also notable are virtual and blurred backgrounds, which Michels said was a game-changer in the early days of the pandemic as people newly working from home worried about revealing more about themselves than they wanted. Employee privacy issues have been raised by many people inside and outside the company, and the ability to add virtual backgrounds or use background blur is an easy way to add privacy, he said. At this point, many cloud video providers add or support virtual or blurred backgrounds, and Microsoft and Zoom are taking this idea a step further. together mode And the latter immersive viewMichels points out. Together mode and immersive view both give users the option to meet in a shared virtual background.
“Captions and translations are very important from an inclusion perspective,” Michels says. He said users with hearing impairments can read along with real-time captions, and real-time translation reduces language barriers.
One company that sees value in real-time translation is toy manufacturer spin masterTyler Pollard, the company's head of digital collaboration, explained this in an interview with No Jitter. Like many other companies, Spin Master had to adopt new ways to collaborate during the pandemic and turned to Zoom's cloud-based video service. Not all employees are native speakers of English, so real-time translation capabilities can help ensure employees can participate in meetings in their native language. Pollard said Spin Master is also exploring additional ways to leverage real-time translation in town hall meetings.
In addition, Spin Master users utilize AI-generated transcripts, Pollard added. But when it comes to video AI, he acknowledged, “There's a fine line we're walking right now.” While out-of-the-box features such as real-time translation and transcription currently provide value to Spin Master, Pollard raised concerns about the capabilities and security of cloud AI, and how providers store and use data.
Videoconferencing AI: The sky is the limit
Cloud video currently has a wealth of AI-based conferencing capabilities, but as Michels pointed out, we are far from the peak of our potential. For example, calendar and schedule management, which is currently done almost entirely manually, is ripe for AI, he added. “Why can't my technology recognize that I'm late for a meeting, tell people, and automatically reschedule it in the background?” he asked, sharing his frustration.
Ultimately, just as the Nest thermostat “observes your life and tries to react to it,” Cloud Video's AI could one day be able to do the same, looking at past behavior and acting accordingly, Weinstein said. He added that AI should be able to resolve small annoyances during work hours, such as making sure your Do Not Disturb setting isn't accidentally left on or lowering the volume of an incoming call if you're in a meeting. “AI needs to serve us, save us time, reduce stress, and prevent mistakes. It doesn't think for us.” [but] They do the work for us,” Weinstein said.
As cloud video providers inch toward the pinnacle of AI, conferencing services will become more natural and intuitive, and “much more inclusive, giving everyone a better seat at the table,” Michels said.
The danger for companies is not to get lost in an avalanche of features. Rather than worrying about the latest and greatest, Weinstein advised focusing on the AI capabilities that you know will bring value to your company. “It would be great if the system would email you a transcription after every meeting, unless you don't need or want a transcription, in which case it doesn't make sense,” he added.
There's no doubt that this message resonated with Spin Master, and Pollard said the company is still exploring how to best use more advanced AI capabilities beyond real-time transcription. “AI definitely brings a lot of value to Spin Master, but I think it's definitely something we really need…to investigate and really understand how AI works in its entirety.”
