Skip to content
HearingTracker
Log in

Open source hearing aid

This topic is closed.

16 replies8.9k viewsLast reply

Link to this topic

Hello. I’d like to draw the attention of the community to the BioAid research project. We have developed a novel sound-processing algorithm for hearing aids based on our knowledge of the signal processing that occurs in the human auditory system.

To get the algorithm out of the laboratory and reach a larger audience, we recently released an implementation of the algorithm in the form of an iOS app. This will turn any recent iOS device into a basic hearing aid. The app is completely free, has no restricted functionality, and contains no advertising. This is a completely non-profit effort. The algorithm source code is also freely available for the curious.

Due to limitations with the hardware, it is certainly not going to replace modern high-end commercially available hearing aids in many situations. However, we hope that the device will benefit some people in certain situations. This is a research tool, so we’d love to hear any feedback, positive or otherwise.

I’m new here, so can’t go about posing links yet, but a web search for bioaid will bring up a website with plenty of information. Thanks for looking

8 people in this conversation6 min readJump to latest

Interesting, I was wondering why people didn’t just make something the size of a cellphone for much cheaper, or just make a cellphone with hearing aids installed. In the age of ipods and cellphones, nobody would ever care… 700$ cellphones vs 5000$++ hearing aids…

cell phone with some bluetooth earphones, could be “new” apple .

However, this would require lots of batteries or constant charger, hand crank anyone? :slight_smile:

If I had moderate loss I would definitely try this though. I think this is the way of the future.

I was wondering why people didn’t just make something the size of a cellphone for much cheaper

That is what we were thinking. Of course, you have to carry the extra device around with you, and there are limitations compared with dedicated hearing aids, but it is a great way to get stuff out of the lab and into the hands of users. Things will only get better with technological innovations in mobile device design and wireless technologies.

If I had moderate loss I would definitely try this though. I think this is the way of the future.

Thanks for the kind comments.

Very interesting idea. Are you working on an android version as well?

A lot of people have started asking me this. The answer is, it depends how much interest there is in the iOS version. It is not long since the release, but interest is slowly mounting.

I decided to target iOS from the outset, because at the time, I knew it was possible to do low-latency real-time audio processing on that platform. Android devices were lagging a bit in that respect back when I started development (although the state of things may be different now). There was also less to worry about in terms of hardware variation, because the apple product lineup is small compared to all the devices that run Android. Even so, the code relatively portable, so Android version is a possibility. It’s just not on my current TODO list.

@nrclark said:

I decided to target iOS from the outset, because at the time, I knew it was possible to do low-latency real-time audio processing on that platform. Android devices were lagging a bit in that respect back when I started development (although the state of things may be different now). There was also less to worry about in terms of hardware variation, because the apple product lineup is small compared to all the devices that run Android. Even so, the code relatively portable, so Android version is a possibility. It’s just not on my current TODO list.

 Most Android users don't realize that few (no?) products run solely on the Google/Android platform.  Samsung, HTC etc. all overlay the "base" Android OS with their own "platform".  OTH, other developers write Android apps that run on virtually all "Android" phones. I would guess this would be more tricky then most apps.

Of necessity I looked into the “settings” of my MP3 player Cowon S9 (now discontinued) and found that volume could be adjusted Left relative to Right + or -. In addition frequency response can be easily adjusted based on 4 freq bands as opposed to Bass, Classical, Pop, Rock and so on.

I do not know how common these features are, this is the first time I became aware of them.

Great App. I would try it if i had the iphone, unfortunately I get an android phone.

I need to you some questions:
which is the minimum (older) version of iphone to run the app?
could you process a voice incoming call on the iphone?
have you evaluated device implementation?

Which is the minimum (older) version of iphone to run the app?

It is compatible with anything running iOS 6 or greater. You can do a web search for “apple ios 6 compatible devices” and you should find pages that will give you the current list.

could you process a voice incoming call on the iphone?

Sadly not. There is no way to access the voice call audio stream for standard phone calls at the moment.

have you evaluated device implementation?

I’m sorry. I’m a little confused by this question. Are you asking if we have looked at implementing the algorithm on a dedicated DSP chip, or if we have evaluated the signal processing in the lab with listeners?

@nrclark said:

I’m sorry. I’m a little confused by this question. Are you asking if we have looked at implementing the algorithm on a dedicated DSP chip, or if we have evaluated the signal processing in the lab with listeners?

I’m asking you if we have looked at implementing the algorithm on a dedicated chip.

(sorry for my bad english)

I downloaded the program on my iPhone. It does not seem to be possible to enter an or to do in-situ fitting. I don’t think pre-programmed responses will help many people. Hearing aid fitting is all about precision in fitting amplification to the hearing loss, with 1-3 precision. I will try the program anyway and give feedback soon.

I’m asking you if we have looked at implementing the algorithm on a dedicated chip

We made a prototype in the lab with a hearing aid connected to a computer setup for the signal processing. This works well as an R&D device. Making a full portable prototype would be a rather tricky process. This is why a smartphone is an attractive device for early testing. It is also a good vehicle for communicating our research work with a wide audience. The lab based work continues, and there are plans to release the data in academic journals in due course. If a manufacturer is interested in the properties of the unique processing in the aid, then they may well experiment with the BioAid concept.

Despite deafness I use telephone every day without aids and BioAid could improve my speech recognition

The feature is not currently available due to limitations with the platform. However, that’s not to say that it would not be possible to incorporate the processing into voice over internet protocol (VOIP) telephony as some kind of audio plugin. Watch this space!

It does not seem to be possible to enter an or to do in-situ fitting. I don’t think pre-programmed responses will help many people.

The self fitting method is certainly a break from tradition. Opinions we’ve heard so far are certainly mixed. Incorporating a test battery into the app is something we’ve been talking about. We’ve been having some success with categorical loudness scaling experiments in the lab, and this may well form the basis of the fitting procedure for BioAid.

In reply to @nrclark

@nrclark said:

The feature is not currently available due to limitations with the platform. However, that’s not to say that it would not be possible to incorporate the processing into voice over internet protocol (VOIP) telephony as some kind of audio plugin. Watch this space!

Yes, you can do it. Use a voip client could be a solution, but the sound quality over the web is not always very very good. Unfortunately compare sounds is not easy.

@nrclark said:

This works well as an R&D device. Making a full portable prototype would be a rather tricky process.

Yes, I understand it. Assembly or c code are not the same on different chipset. I hope to see low cost and strong aids in the future, and opensource algorithms could contribute it, though unfortunately hardware is a really issue

Do you ever develop with audio API (or evaluated it)?

Related discussions