en de

Online Magazine

With AI against background noise

We are all familiar with the "cocktail party effect" – you can no longer understand your interlocutor because of all the background noise. When artificial intelligence filters out these noises, it's not just party-goers who benefit, but also content creators and people with hearing difficulties. The Graz-based start-up clir.ai makes this possible with an app.

by Tobias Imbach

This episode of our podcast series Cat!apult is once again about an exciting idea that uses the potential of data and artificial intelligence to improve or simplify our daily lives.

CEO Andreas Krassnitzer and his start-up clir Technologies from Graz have set themselves the goal of reducing background noise in real time using AI-based algorithms.

In the podcast, he talks about how clir.ai not only enables in-depth conversations in noisy bars or podcast recordings with the smartphone, but also how people with hearing aids benefit from the noise filter.


  • We’re using deep neural networks to distinguish between speech and anything that is not speech in your surroundings.
  • To train our AI, we dynamically mix speech and noise data, and thus the AI learns what is speech and what isn't.
  • It’s a very similar challenge to what I’ve experienced when working in autonomous driving – you can distinguish between traffic signs and the background, or you can recognize voice in contrast to other sounds.
  • Our AI enables headphones and hearing aids to do in an enhanced way what our brain is doing: focus on a specific sound source and blend out the rest.
  • Even expensive hearing aids can’t help you in differentiating individual sound sources. The training that people would have to do to learn that again would be too difficult for anyone, and that’s where AI comes in.
  • Currently, part of the sound quality deteriorates during the process, but this problem will be solved with larger networks and more computing power.
  • Once we've isolated the individual sound sources, we will be able to do things like live translation and improve speech to text functions.

Did you know that start-ups also use AI to ...

... help businesses save energy?
Listen to the podcast.

... prevent dementia?
Join the conversation with HILDA CEO Thomas Marek.

... help visually impaired people find their way in new environments?
Find out more in this podcast.

Listen to this podcast on another platform:




Andreas Krassnitzer

clir.ai, Graz

Andreas Krassnitzer has experienced first-hand the challenges faced by people with hearing impairment: he grew up side by side with an older brother who suffers from severe hearing loss. While working on artificial intelligence for autonomous driver assistance systems, he realised that AI and IoT technologies could also change the way hearing aids work, and thus the lives of people with hearing impairment. Andreas pursued this idea further when he founded the company clir Technologies together with Stefan Stücklschweiger in 2021.


Data analytics Machine learning

7 habits to shorten the time-to-value in process mining
AI in medicine Data analytics

The smart drinking cup
AI in business Data analytics Machine learning

How can banks become truly AI-driven?