BabbleLabs is now part of Cisco

The Case for No-Touch Huddle Rooms

Enterprises designed and built an estimated 32.4 million huddle rooms worldwide that host audio, video and display system technologies to foster collaboration. While we can presume that most of these rooms have been sitting empty during the pandemic, you can bet they will become some of the most important places in any office when employees return to work.

In an effort to decrease the risk of spreading COVID-19, remote communications will continue to be a critical part of business. Limitations on how many people can be in an office will require employees to continue collaborating with each other from multiple locations – some working from the office and some working from home. And restrictions on non-essential travel and personal preferences will dramatically decrease the number of face-to-face meetings with partners, prospects and customers. There will be a far greater number of virtual meetings and a bigger reliance on unified communications (UC) technologies.

But with warnings about the possible risk of contracting coronavirus from affected surfaces , many employees will be uncomfortable sharing rooms, tools and equipment. To address some concerns, experts recommend organizations expand or move huddle rooms to larger meeting spaces to meet social distancing standards and reduce touch points. In fact, it’s entirely possible and quite simple to remove all touch points. Installing AI-based speech recognition software like BabbleLabs Clear Command can reduce the risk of spreading disease while providing additional benefits such as eliminating the need for UC-related IT support and providing ...

Continue Reading

Human-in-the-loop approach for AI-based speech enhancement assessments

In the age of remote collaboration and videoconferencing, AI plays an important role in speech processing to remove background noise, enhance speech, and gain insights into speech and audio streams. But how do you assess the efficacy of such technology in the context of decades of legacy hardware and software solutions? At BabbleLabs, in addition to objective metrics, we use a listener perception-based framework to evaluate the efficacy of our AI-based speech enhancement models and products.

Subjective assessments with listener perceptions on sound quality

The gold standard for assessing speech and sound quality utilizes the subjective opinions of a large, diverse panel of human listeners. Traditionally, this process is laborious and expensive to conduct. There are established objective measures using predictive models. However, the predictions may be valid only for very specific types of distortions, and useful, at the earlier stages of audio algorithm development.

At BabbleLabs, we use objective measures (such as: PESQ, ESTOI, SNR) in the early stages of the algorithm development, since they are fast and inexpensive to apply. In the intermediate and, especially at the final stages, we place greater reliance on subjective, human listener opinions in order to:

Leverage the gold standard in sound quality assessment Get feedback from real listeners in real-life listening environments Gather reliable opinions for many types of audio distortions

To this end, we have developed a simple, comparative, crowdsourced, subjective testing framework that is performed at scale with a large number of individuals in “our ...

Continue Reading