All Speech and Audio Processing Group News

Idiap researchers are Lemanic Life Science Hackathon winners
by Maxime Bourlard — last modified Jun 19, 2024 03:08 PM

Tilak Purohit and Barbara Ruvolo, contributing to Idiap’s “AI for Life” research program, participated and won 1st prize at the Lemanic Life Science Hackathon organized at EPFL at the end of April 2024. Their winning project resulted in an AI-enabled user interface prototype to support the detection of depression via speech.

Idiap awards 2022
by nicolas-filippov — last modified Dec 21, 2022 11:40 AM

Every year, the Institute nominates two students for its internal awards. In 2022, the Paper Award goes to Alexandre Bittar, and the Student Award goes to Teguh Lembono. Congratulations!

Artificial intelligence to answer challenges of Mexican society
by Maxime Bourlard — last modified Nov 25, 2022 02:01 PM

Esau Villatoro, research associate at the Speech & Audio Processing group at Idiap, and his colleagues from the Mathematics Research Center (CIMAT) from Mexico have won first place in two competitions related to Natural Language Processing. The objective of these competitions is to improve important aspects of Mexican society such as tourism and communication.

A speech processing system inspired by the human brain
by nicolas-filippov — last modified Oct 19, 2022 02:48 PM

Idiap researchers published a paper describing an approach to speech processing based on the properties of the human brain. Their method proved as efficient as the current standard, whilst conserving the advantage of energy efficiency. Moreover, their work is replicable thanks to open access software paving the way for future applications.

No artificial intelligence without full mastery of maths and statistics
by nicolas-filippov — last modified Jul 27, 2022 05:27 PM

Neural networks are often among cited technologies when it comes to artificial intelligence latest exploits. The downside is that this technology can be demanding in terms of energy costs. Two Idiap researchers demonstrated that in certain cases you should rather go for classical maths rather than for the AI hype.

How applied research impacted his scientific career
by Maxime Bourlard — last modified Mar 01, 2022 06:01 PM

Arriving in 2019 for a sabbatical year from the University of Mexico, Esaú Villatoro has now been working at Idiap for more than two years. Between publishing his work and adapting to Swiss life, he looks back on his experience at the institute.

Idiap is looking for its next director
by nicolas-filippov — last modified Jan 19, 2022 05:13 PM

Idiap Research Institute and the School of Engineering at EPFL invite applications for the directorship of Idiap. The successful candidate will also hold a faculty position as full professor at EPFL School of Engineering.

Idiap awards 2021
by nicolas-filippov — last modified Dec 21, 2021 04:47 PM

The Institute nominates every year two students for its internal awards. In 2021, the Best Paper Award goes to Suhan Shetty, and the Best Student Award goes to Parvaneh Janbakhshi. Congratulations!

Over 12 millions Swiss francs for more inclusive ICT
by nicolas-filippov — last modified Nov 03, 2021 01:39 PM

Access to information is a challenge for disabled people, even at a time when communications channels are increasing. An international consortium gathering researchers, as well as private and public partners, under the leadership of the University of Zurich and including Idiap and Icare from the French speaking side of Switzerland was granted 6 million Swiss francs from Innosuisse—completed by 6 million from private partners—to take up this challenge.

Revisiting autoencoders for speech representation learning
by admin — last modified Jul 16, 2021 02:16 PM

“The deep artificial neural networks are inspired from the hierarchical structure of the human brain. Therefore, questioning the neural networks to learn robust, generalizable representations (abstractions) just like the human brain is not absurd, but challenging. However, revisiting fundamentals can be helpful.”

Prof. Hervé Bourlard to receive the Prestigious IEEE 2022 James L. Flanagan Speech and Audio Processing Award
by nicolas-filippov — last modified Jul 07, 2021 08:49 AM

Director of the Idiap Research Institute and Professor at EPFL, Hervé Bourlard is recognized for his major contributions to neural networks for statistical speech recognition. This distinction is awarded to him jointly with his colleague and long-time friend, Prof. Nelson Morgan of the International Computer Science Institute, and the University of California at Berkeley.

Idiap at ICASSP 2021
by admin — last modified Apr 21, 2021 11:20 AM

Idiap's speech group will be presenting six papers at ICASSP 2021. Our papers address a variety of research problems in pathological and physiological speech processing, automatic speech recognition, and machine learning.

Estimating Breathing Pattern and Parameters from Speech Waveform
by admin — last modified Mar 25, 2021 05:19 PM

Alternation in respiratory system and speech production system results in changes in speech. Therefore, speech signal, which can be acquired in a non-invasive manner, could be used to predict breathing patterns. There is a growing interest in that direction, which has gained further momentum with COVID-19 situation.

HAAWAII H2020 EC project
by admin — last modified Feb 08, 2021 04:57 PM

Highly Automated Air Traffic Controller Workstations with Artificial Intelligence Integration

HMMGradients.jl
by admin — last modified Jan 29, 2021 08:49 AM

A first step towards a Julia-based speech recognition toolkit.

Idiap at NeurIPS 2020
by admin — last modified Dec 17, 2020 11:21 AM

This week Idiap's speech and machine learning group presented their joint work on Fast Transformers with Clustered Attention

Lei Wang finishes her internship at Idiap
by admin — last modified Dec 02, 2020 04:08 PM

Lei joined Idiap in October 2019 for an internship in the framework of the China Scholarship Council visiting scholar program. During her internship, she worked on neural network-based mappings for single-channel dereverberation and noise reduction.

The first Roxanne field test took place
by admin — last modified Nov 02, 2020 05:39 PM

Roxanne is a EU-funded project that leverages text, speech and video in real-time in order to build tools for combating organized crime.