This post accompanies our presentation “Immersive multitalker remote microphone system” at the 181st Acoustical Society of America Meeting in Seattle.
In our previous post, which accompanied a paper at WASPAA 2021, we proposed an improved wireless microphone system for hearing aids and other listening devices. Unlike conventional remote microphones, the proposed system works with multiple talkers at once, and it uses earpiece microphones to preserve the spatial cues that humans use to localize and separate sound. In that paper, we successfully demonstrated the adaptive filtering system in an offline laboratory experiment.
To see if it would work in a real-time, real-world listening system, we participated in an Acoustical Society of America hackathon using the open-source Tympan platform. The Tympan is an Arduino-based hearing aid development kit. It comes with high-quality audio hardware, a built-in rechargeable battery, a user-friendly Android app, a memory card for recording, and a comprehensive modular software library. Using the Tympan, we were able to quickly demonstrate our adaptive binaural filtering system in real hardware.
The Tympan processor connects to a stereo wireless microphone system and binaural earbuds.
This post accompanies the paper “Adaptive Binaural Filtering for a Multiple-Talker Listening System Using Remote and On-Ear Microphones” presented at WASPAA 2021 (PDF).
Wireless assistive listening technology
Hearing aids and other listening devices can help people to hear better by amplifying quiet sounds. But amplification alone is not enough in loud environments like restaurants, where the sound from a conversation partner is buried in background noise, or when the talker is far away, like in a large classroom or a theater. To make sound easier to understand, we need to bring the sound source closer to the listener. While we often cannot physically move the talker, we can do the next best thing by placing a microphone on them.
Remote microphones make it easier to hear by transmitting sound directly from a talker to a listener. Conventional remote microphones only work with one talker at a time.
When a remote microphone is placed on or close to a talker, it captures speech with lower noise than the microphones built into hearing aid earpieces. The sound also has less reverberation since it does not bounce around the room before reaching the listener. In clinical studies, remote microphones have been shown to consistently improve speech understanding in noisy environments. In our interviews of hearing technology users, we found that people who use remote microphones love them – but with the exception of K-12 schools, where remote microphones are often legally required accommodations, very few people bother to use them.
Nearly every modern hearing aid uses an algorithm called dynamic range compression (DRC), which automatically adjusts the amplification of the hearing aid to make quiet sounds louder and loud sounds quieter. Although compression is one of the most important features of hearing aids, it might also be one of the reasons that they work so poorly in noisy environments. Hearing researchers have long known that when DRC is applied to multiple sounds at once, it can cause distortion and make background noise worse. Our research team is applying signal processing theory to understand why compression works poorly in noise and exploring new strategies for controlling loudness in noisy environments.
Although source separation (separating distinct and overlapping sound sources from each other and from dispersed noise) in a small, quiet lab with only a few speakers usually produces excellent results, such a situation may not always be present. In a large reverberant room with many speakers, for example, it may be difficult for a person or speech recognition system to keep track of and to comprehend what one particular speaker is saying. But using source separation in such a scenario to improve intelligibility is quite difficult without having external information that in itself may also be difficult to obtain.
Many source separation methods work well up to only a certain number of speakers – typically not much more than four or five. Moreover, some of these methods rely on constraining the number of microphones to an amount equal to the number of sources, and will scale poorly in terms of results, if at all, with the addition of more microphones. Limiting the number of microphones to the number of speakers will not work in these difficult scenarios, but adding more microphones may help, due to the greater amount of spatial information made available by the additional microphones. Therefore, the motivation behind this experiment was to find a suitable source separation method, or series of such methods, that could leverage the “massive” number of microphones used in the Massive Distributed Microphone Array Dataset and solve the particularly challenging problem of separating ten speech sources. Ideally, the algorithm would rely on as little external information as possible, instead relying on the wealth of information gathered by the microphone arrays distributed around the conference room. Thus, the delay-and-sum beamformer was first considered for this task, because it only requires the locations of each source and microphone, and it inherently scales well with a large number of microphones.
Shown above is a diagram depicting the setup for the Massive Distributed Microphone Array Dataset. Of note is the fact that there are two distinct types of arrays – wearable arrays, denoted by the letter W and numbered from 1-4, and tabletop arrays, denoted by the letter T and numbered from 1-12. Wearable arrays have 16 microphones each, whereas tabletop arrays have 8.
Face masks are a critical tool in slowing the spread of COVID-19, but they also make communication more difficult, especially for people with hearing loss. Face masks muffle high-frequency speech sounds and block visual cues. Masks may be especially frustrating for teachers, who will be expected to wear them while teaching in person this fall. Fortunately, not all masks affect sound in the same way, and some masks are better than others. Our research team measured several face masks in the Illinois Augmented Listening Laboratory to find out which are the best for sound transmission, and to see whether amplification technology can help.
Several months into the pandemic, we now have access to a wide variety of face masks, including disposable medical masks and washable cloth masks in different shapes and fabrics. A recent trend is cloth masks with clear windows, which let listeners see the talker’s lips and facial expressions. In this study, we tested a surgical mask, N95 and KN95 respirators, several types of opaque cloth mask, two cloth masks with clear windows, and a plastic face shield.
We measured the masks in two ways. First, we used a head-shaped loudspeaker designed by recent Industrial Design graduate Uriah Jones to measure sound transmission through the masks. A microphone was placed at head height about six feet away, to simulate a “social-distancing” listener. The loudspeaker was rotated on a turntable to measure the directional effects of the mask. Second, we recorded a human talker wearing each mask, which provides more realistic but less consistent data. The human talker wore extra microphones on his lapel, cheek, forehead, and in front of his mouth to test the effects of masks on sound capture systems.
Microphone arrays are important tools for spatial sound processing. Traditionally, most methods for spatial sound capture can be classified as either beamforming, which tries to isolate a sound coming from a single direction, or source separation, which tries to split a recording of several sounds into its component parts. Beamforming and source separation are useful in crowded, noisy environments with many sound sources, and are widely used in speech recognition systems and teleconferencing systems.
Since microphone arrays are so useful in noisy environments, we would expect them to work well in hearing aids and other augmented listening applications. Researchers have been building microphone-array hearing aids for more than 30 years, and laboratory experiments have consistently shown that they can reduce noise and improve intelligibility, but there has never been a commercially successful listening device with a powerful microphone array. Why not?
The problem may be that most researchers have approached listening devices as if they were speech recognition or teleconferencing systems, designing beamformers that try to isolate a single sound and remove all the others. They promise to let the listener hear the person across from them in a crowded restaurant and silence everyone else. But unlike computers, humans are used to hearing multiple sounds at once, and our brains can do a good job separating sound sources on their own. Imagine seeing everyone’s lips move but not hearing any sound! If a listening device tries to focus on only one sound, it can seem unnatural to the listener and introduce distortion that makes it harder, not easier, to hear.
A source remixing system changes the relative levels of sounds in a mixture while preserving their spatial cues.
This paper proposes a new type of array processing for listening devices: source remixing. Instead of trying to isolate or separate sound sources, the system tries to change their relative levels in a way that sounds natural to the listener. In a good remixing system, it will seem as if real-world sounds are louder or quieter than before.
When our team designs wearable microphone arrays, we usually test them on our beloved mannequin test subject, Mike A. Ray. With Mike’s help, we’ve shown that large wearable microphone arrays can perform much better than conventional earpieces and headsets for augmented listening applications, such as noise reduction in hearing aids. Mannequin experiments are useful because, unlike a human, Mike doesn’t need to be paid, doesn’t need to sign any paperwork, and doesn’t mind having things duct-taped to his head. There is one major difference between mannequin and human subjects, however: humans move. In our recent paper at WASPAA 2019, which won a best student paper award, we described the effects of this motion on microphone arrays and proposed several ways to address it.
Beamformers, which use spatial information to separate and enhance sounds from different directions, rely on precise distances between microphones. (We don’t actually measure those distances directly; we measure relative time delays between signals at the different microphones, which depend on distances.) When a human user turns their head – as humans do constantly and subconsciously while listening – the microphones near the ears move relative to the microphones on the lower body. The distances between microphones therefore change frequently.
In a deformable microphone array, microphones can move relative to each other.
Microphone array researchers have studied motion before, but it is usually the sound source that moves relative to the entire array. For example, a talker might walk around the room. That problem, while challenging, is easier to deal with: we just need to track the direction of the user. Deformation of the array itself – that is, relative motion between microphones – is more difficult because there are more moving parts and the changing shape of the array has complicated effects on the signals. In this paper, we mathematically analyzed the effects of deformation on beamformer performance and considered several ways to compensate for it.
Imagine what it would sound like to listen through someone else’s ears. I don’t mean that in a metaphorical sense. What if you had a headset that would let you listen through microphones in the ears of someone else in the room, so that you can hear what they hear? Better yet, what if your headset was connected to the ears of everyone else in the room? In our group’s latest paper, “Cooperative Audio Source Separation and Enhancement Using Distributed Microphone Arrays and Wearable Devices,” presented this week at CAMSAP 2019, we designed a system to do just that.
Our team is trying to improve the performance of hearing aids and other augmented listening devices in crowded, noisy environments. In spaces like restaurants and bars where there are many people talking at once, it can be difficult for even normal-hearing people to hold a conversation. Microphone arrays can help by spatially separating sounds, so that each user can hear what they want to hear and turn off the sounds they don’t want to hear. To do that in a very noisy room, however, we need a large number of microphones that cover a large area.
Complex listening environments include many different sound sources, but also many microphone-equipped devices. Each listening device tries to enhance a different sound source.
In the past, we’ve built large wearable microphone arrays with sensors that cover wearable accessories or even the entire body. These arrays can perform much better than conventional earpieces, but they aren’t enough in the most challenging environments. In a large, reverberant room packed with noisy people, we need microphones spread all over the room. Instead of having a compact microphone array surrounded by sound sources, we should have microphones spread around and among the sound sources, helping each listener to distinguish even faraway sounds.
This post describes our new massive distributed microphone array dataset, which is available for download from the Illinois Databank and is featured an upcoming paper at CAMSAP 2019.
The conference room used for the massive distributed array dataset.
Listening in loud noise is hard: we only have two ears, after all, but a crowded party might have dozens or even hundreds of people talking at once. Our ears are hopelessly outnumbered! Augmented listening devices, however, are not limited by physiology: they could use hundreds of microphones spread all across a room to make sense of the jumble of sounds.
Our world is already filled with microphones. There are multiple microphones in every smartphone, laptop, smart speaker, conferencing system, and hearing aid. As microphone technology and wireless networks improve, it will be possible to place hundreds of microphones throughout crowded spaces to help us hear better. Massive-scale distributed arrays are more useful than compact arrays because they are spread around and among the sound sources. One user’s listening device might have trouble distinguishing between two voices on the other side of the room, but wearable microphones on those talkers can provide excellent information about their speech signals.
Many researchers, including our team, are developing algorithms that can harness information from massive-scale arrays, but there is little publicly available data suitable for source separation and audio enhancement research at such a large scale. To facilitate this research, we have released a new dataset with 10 speech sources and 160 microphones in a large, reverberant conference room.