logo
episode-header-image
May 2021
38m 43s

The first African voice assistant

Bbc World Service
About this episode

Speech smart assistants currently do not support any African language, but now Mozilla’s Common Voice project is building a dataset for Kiswahili which is spoken by more than a 100 million people in Kenya, Uganda, Tanzania, Rwanda, Burundi, and South Sudan. They have just been awarded almost $5m for the project. Remy the community lead at Common Voice Kinyarwanda and Chenai chair special adviser for Africa Innovation at the Mozilla Foundation tells us more about the work.

Federated Learning As more of our data is processed through machine learning systems, Dr Nic Lane, of the Cambridge Computer Laboratory, explains a new way of processing and using our data without the need to send it through data centres. One solution to reduce the impact is federated learning – where data is processed on the edge, on your device, rather than being centralised. Not only does this reduce the environmental impact of computing, but it also has benefits for privacy and opens up opportunities for data sharing between companies.

Electronic VR socks Reporter Claire Jordan has been investigating a novel approach to “walking” in a virtual reality. Existing ways are expensive and need considerable space to function, as well as having a less than satisfactory user experience. A new technology being developed by Professor Yusuke Matsuda of Toyohashi University of Technology in Japan could expand the reach of immersive VR movement to a much wider audience and deliver a better user experience than current solutions. The user wears electronic socks that create the feeling of walking in VR even though the user is sitting down. This could allow users to spend much longer in VR and also allow people with mobility issues to move around more freely, and possibly feel as though they are moving, in VR.

The programme is presented by Gareth Mitchell with expert commentary from Bill Thompson. Producer: Ania Lichtarowicz

(Image: Mozilla’s Common Voice project. Credit: Mozilla) (Image: Mozilla’s Common Voice project. Credit: Mozilla)

Up next
May 2023
Tech Life
How is technology changing our lives? Search for Tech Life wherever you get your podcasts and subscribe or follow if you can. 
34s
Mar 2023
Digital Planet says goodbye
On Digital Planet’s final ever show we discuss the legacy of Gordon Moore, the father of transistors and creator of Moore’s law.Special guests this week are Angelica Mari and Ghislaine Boddington.The programme is presented by Gareth Mitchell with expert commentary from Bill Thomp ... Show More
52m 16s
Mar 2023
3D printed food – what’s cooking?
Could 3D-printing be serving us up a tasty food revolution, or is it the ultimate in gimmicky processed foods taking us yet further away from natural eating? In the kitchen, a 3D-printer builds up customised tasty treats like exotic cheesecakes, layer by layer, using edible paste ... Show More
52m 5s
Recommended Episodes
Feb 2022
🌍 AI in Africa - Voice & language tools (Practical AI #167)
In the third of the “AI in Africa” spotlight episodes, we welcome Kathleen Siminyu, who is building Kiswahili voice tools at Mozilla. We had a great discussion with Kathleen about creating more diverse voice and language datasets, involving local language communities in NLP work, ... Show More
43m 37s
Feb 2022
Voice for Everyone with Hillary Juma from Mozilla Common Voice
<p>Common Voice is a crowdsourcing project started by Mozilla to create a free database for speech recognition software. The project is supported by volunteers who record sample sentences with a microphone and review recordings of other users. Community Manager Hillary Juma talks ... Show More
32m 20s
Mar 2020
Speech recognition to say it just right
Catherine Breslin of Cobalt joins Daniel and Chris to do a deep dive on speech recognition. She also discusses how the technology is integrated into virtual assistants (like Alexa) and is used in other non-assistant contexts (like transcription and captioning). Along the way, she ... Show More
49m 15s
Mar 2020
Speech recognition to say it just right (Practical AI #82)
Catherine Breslin of Cobalt joins Daniel and Chris to do a deep dive on speech recognition. She also discusses how the technology is integrated into virtual assistants (like Alexa) and is used in other non-assistant contexts (like transcription and captioning). Along the way, she ... Show More
49m 14s
Jul 2023
2451: How LingQ is Expanding Language Horizons With Tech
In today's episode of Tech Talks Daily, we are joined by Mark Kaufmann, the CEO of LingQ, an innovative language-learning app. LingQ has recently made headlines for announcing the addition of Swahili to their extensive catalogue of languages, raising their total to 42 languages o ... Show More
30m 21s
Mar 2025
How AI can bridge the Deaf and hearing worlds | Adam Munder
Software engineer Adam Munder is on a mission to break down communication barriers between the Deaf and hearing worlds. In a live demo, he introduces OmniBridge — an AI platform that translates American Sign Language into English text in real time — and demonstrates how this tech ... Show More
6m 44s