Using language technology to assist the hard of hearing

The Nordic association of the hard of hearing (Nordiska Hörselskadades Samarbetskommitté, NHS) had a seminar at Hotel Selfoss last week. On Friday, Anna gave a talk there on how language technology might assist people hard of hearing to communicate and access information in a predominately hearing world. Automatic transcription of live communication and automatic caption of video material is already working for English and some other languages, and the Nordic participants of the seminar were eager to see this technology advance in their languages. At LVL, we are working on open ASR systems, making the development of technology like this possible for Icelandic.

The rest of the slides can be viewed by selecting the first slide below.

NHS_cover_icon

Meeting with Mycroft

This week LVL sat down with Mycroft to discuss the possibilities of collaborating and working together to bring more speech technology to Iceland.  We discussed using Mozilla’s Common Voice to bring about another open source Icelandic speech dataset, and possibly an Icelandic voice assistant. The Mozilla project requires just 5,000 phrases which anyone can contribute, even you!

mycroft_at_RU
LVL meets with Joshua Montgomerey of Mycroft

RANNIS Infrastructure grant 2018 – Awarded.

The complexity of neural network models increases every year and it takes a lot to keep up with computational hardware fast enough to train them efficiently.  Earlier this year, we applied to the RANNIS Infrastructure Fund for funding to expand our current HPC cluster.  We are happy that our proposal “Deep Learning Infrastructure for Speech and Language Technology” was selected to be funded. Only three grants were granted to Reykjavik University, and ours was one of them. This money will allow us to buy a fully equipped SuperMicro 4028GR-TR2 server with NVIDIA 1080Ti GPUs. We hope to sign the grant contract within several weeks and then order the machine. Next comes the process of assembling and integrating it into our current cluster. I can say that several group members can’t wait to have more power available to them.

 

02I72646
List of awarded grants – RANNIS

First conference of the year is LREC!

This year, Jón will be representing our LVL group at the LREC 2018 conference happening this week in gorgeous Miyazaki, Japan. Jón will be presenting the paper,  “Open ASR for Icelandic: Resources and a Baseline System.” As co-author, Anna, says, “The paper describes the language resources used in the project Open ASR for Icelandic: the Málrómur speech corpus, the Leipzig Corpora Collection and the Icelandic pronunciation dictionary, and their processing for the utilization in the training of the ASR system. Furthermore, we experiment with different content of the acoustic training corpus to examine the impact of carefully selected speech data on the WER of the ASR system.” To learn more, read their paper and visit ASR_Resources_LREC2018_A0_portrait_final to get the details that they weren’t able to fit in.

We wish Jón a good conference and lots of fun on a different island!

For those who weren’t able to view the poster in person, we have a PDF of it below:

LRECpaper

Open Icelandic ASR unveiled at the Language Technology Seminar

On the 27th of April, we will be presenting the web portal for our project, a Free and Open Speech Recognition for Icelandic. The ASR will be the opening topic of the Language Technology Seminar, followed by other language technology talks and presentations. The seminar takes place at Reykjavík University, room M101 on Friday, April 27th, 2018 12:00PM.

More detailed information can be found at the following link: Announcing the New Open ASR for Icelandic and the Facebook event.

The automatic speech recognizer was previously demoed at UTMessan and University Day but after the seminar it will be publicly available at https://tal.ru.is/ for newer devices. To get a better idea of how our Open Icelandic ASR works, watch the following Icelandic news segment:

But if you don’t know any Icelandic and just want to use the API, then follow us for updates.

ASR: A SMASHING HIT AT UTMESSAN!!!

On Saturday Róbert and Anna also presented a prototype for an automatic speech recognition web service for general Icelandic (Vefgátt fyrir íslenskan talgreini). This speech recognizer for Icelandic will be made available to the public later this spring.

ASRImageShopped
LVL members explain how to use the ASR.
ASRDemoTesters
Róbert presents the microphone to attendees to try out the Icelandic ASR for themselves.

 

Does the future understand Icelandic? – UTMessan Talk

On Friday, LVL will be at the UTMessan conference being held here in Reykjavik. UTMessan is a IT conference presented in both Icelandic and English. The talk, “Skilur framtíðin íslensku? (Does the future understand Icelandic?)” will be presented by Anna Björk. This will be our first time presenting at UTMessan so we are very excited.

If you are interested in attending, below are the talk details:

Abstract: In recent years language technology and artificial intelligence have made a giant leap forward. The revolution becomes clear in the development of intelligent assistants like Siri, Alexa, Google Assistant and Cortana, that in ever more households are becoming a part of the family. In the discussion about language technology for Icelandic, the most prominent subject has been how and when we are going to speak Icelandic with those assistants.

But what about language technology for the purpose of increasing efficiency and service quality of companies and institutions? This talk discusses language technology in this context and what is needed so that Icelandic companies can join the rapid development in the field.

Watch the talkhttps://www.youtube.com/watch?v=IaOYG23R7_k
Date: 2 February, 2018
Location: Tækjatal – Chatbot Talks, Silfurberg B, Harpa, Reykjavík, Iceland
Language: Icelandic (perfect opportunity to brush up on your Icelandic)
Time: 11:55am – 12:25pm
Speaker:

AnnaUTMessanTalk

More details at UTMessan

We hope to see you there!

Interspeech 2017 Conference Summary

Yu-Ren gave a really good talk about measuring voice severity.

Image uploaded from iOS2

Then, on Wednesday Anna, Inga, Matthías, and Jón answered questions about their posters to conference attendees.

Image uploaded from iOS15

During the conference, welcome reception, and the banquet we reconnected with many old colleagues and met many new  in the speech processing and speech recognition world. For anyone who didn’t get a chance to attend Yu-Ren’s talk or to see the posters by Jón and Inga, we have links at the bottom of the post.

This slideshow requires JavaScript.

Anna’s data is also available on Malfong now.

Inga’s Icelandic Parliament ASR Corpus

Yu-Ren’s slides

Jón’s Eyra Speech Corpora Poster

We hope to see you next year!

Largest Icelandic LVL Group at Interspeech in 2017

This August, our LVL group members will be attending Interspeech 2017 to present their three papers, meet other folks in the speech recognition field, and have lots of fun. Two posters,“Building an ASR corpus using Althingi’s Parliamentary Speeches” and “Building ASR corpora using Eyra,” will be presented during the Wednesday Special Session: Digital Revolution for Under-resourced Languages 2 poster session at 13:30-15:30 so go say, “Hi!” to our members if you can. The third paper, “Objective Severity Assessment From Disordered Voice Using Estimated Glottal Airflow,” will be presented as a talk by Yu-Ren on Monday afternoon.

For the Alþingi Speech paper by Inga Rún, the language corpus can be found at Malfong.is and the Kaldi recipe can be found on Github.com but the best resource will be Inga Rún herself so grab a drink and find her at the Welcome Reception or the Standing Banquet.

We hope to meet you all there!