Petr Motlicek - home page#
Open positions at Idiap#
Feel free to contact me anytime by email, or subscribe directly to one of open positions below:
Contact#
Mailto:petr dot motlicek add idiap dot chTel:+41 27 721 77 49Place:Rue Marconi 19, 1920, Martigny, Switzerland
Home page at Faculty of Information Technology, Brno University of Technology, CZ
profile
Biography#
I am a senior researcher at Idiap Research Institute in Martigny, Switzerland, external teacher at EPFL and also an associate professor at Brno University of Technology (BUT), Faculty of Information Technology (FIT). My research interests include audio and speech processing (especially speech and speaker recognition), and its applications for various tasks (i.e. embedded platforms, robotics, forensic voice comparison).
Part of my scientific work was also done at Oregon Graduate Institute, USA (2000-2001) and at Ecole Superieure d’Ingenieurs en Electrotechnique et Electronique, France (1999).
Speech/audio/signal processing
Artificial intelligence
Machine learning
Habilitation for the position of associate proffesor, 2022, Faculty of Information Technology, Brno University of Technology
PhD in Computer Science, 2003, Faculty of Information Technology, Brno University of Technology
Masters in Electrical Engineering, 1999, Faculty of Electrcial Engineering , Brno University of Technology
[EUROCONTROL] industrial tender on integrating speech and text processing technologies into the ATC simulator (2023-2024)
CRiTERIA EC H2020 (2021-2024)
TRACY EC project (topic DIGITAL-2022-DEPLOY-02-LAW-SECURITY-AI) on big-data analytics from base-stations registrations and CDRs e-evidence system (2023-2025)
Armasuisse project on “Speech analytics for communications intelligence” (2022-2023)
Several industrial collaborations are ongoing …
A detailed list of publications can be found at Idiap publications site
List of publications before 2005 is at my FIT website
Zuluaga Gomez, Jaun Pablo, EPFL Lausanne, Switzerland
Nigmatulina, Iuliia, University of Zurich, Szitzerland
Prasad, Amrutha, Brno University of Technology, Czech Republic
(update 11/2022)
Dr. Petr Motlicek received the M.Sc. degree in electrical engineering and the Ph.D. degree in computer science from Brno University of Technology (BUT), Czech Republic, in 1999 and 2003, respectively. In 2000, he conducted research on very low bit-rate speech coding at Ecole Superieure d’Igenieurs en Electrotechnique et Electronique (ESIEE), Paris, France. From 2001 to 2002, he was a research intern at Oregon Graduate Institute (OGI), Portland, U.S., collaborating with Prof. Hermansky in the area of distributed speech recognition. Dr. Motlicek currently holds a position of assistant professor at Faculty of Information Technology, BUT. He is aligned with Department of Computer Graphics and Multimedia led by Prof. Cernocky.
From 2005, Dr. Motlicek has been a Postdoc Researcher at Idiap Research Institute, Martigny, Switzerland, working in the area of low bit-rate speech coding. In 2018, he received a position of senior research scientist at Idiap, focusing especially on R&D activities in the domain of automatic multilingual speech recognition, speaker recognition as well as spoken language understanding. He is currently specialising on security related research (EU FCT R&D), and also in EU research for automation in air-traffic management.
From 2012, Dr. Motlicek is an external lecturer at Ecole Polytechnique Federale, Lausanne, Switzerland, teaching speech processing related courses in Ms.C. and Ph.D. programmes.
In summer 2010, Dr. Motlicek participated in the ASR workshop coordinated by Dr. Daniel Povey resulting in a new speech processing framework called KALDI.
From 2000, Dr. Motlicek has been a member of IEEE (senior member from 2008) and ISCA.
Dr. Motlicek has been the project coordinator of:
He has been participating as researcher or principal investigator in several other European projects such as TA2 and TA2-EEU, DBOX, DIRAC, AMI and AMIDA, DBOX, SiiP, Malorca, MuMMER, Haawaii, CRiTERIA. Dr. Motlicek participated in recently ended IARPA project called SARAL, funded by the U.S. government.
Dr. Motlicek has collaborated with industries such as Qualcomm and Samsung on patent applications.
Madikeri, Srikanth, Research Associate
Villatoro, Esau, Research Associate
Burdisso, Sergio, Research and Development Engineer
Khalil, Driss, Development engineer
Maxime Vidal, Engineer
Tkaczuk, Jakub, Postcoc at Idiap, 2022
Sarfjoo, Saeed, Research and Development Engineer at Idiap, 2017 - 2022
Weipeng He, PhD at Idiap
Subhadeep Dey, PhD at Idiap
Muskaan Singh, Postdoc at Idiap
Shantipriya Parida, postdoc at Idiap
Mael Fabien, PhD at Idiap
Qingran Zhan, intern at Idiap
Ajay Srinivasamurthy, postdoc at Idiap
Ivan Himawan, postdoc at Idiap
Gyorgy Szaszak, postdoc at Idiap
Chloe Salamin, intern at Idiap
Recent (and less recent) posts:#
August 2023 Work on Factual assessment: As part of Horizon EC project CRiTERIA, Idiap team (Burdisso, Sergio,Daira Sanchez and Petr Motlicek) has worked on designing new concept of factual assessment system trying to mimic the competent duo “Dr. Watson” and “Sherlock Holmes“. More info is here. Further details can also be found here.
August 2023: Interspeech 2023 papers presented in Dublin, with list of all papers here. Our presented papers:
August 2023 News from JSALT: IDIAP team (Madikeri, Srikanth and Burdisso, Sergio) has been participating on this years (6 weeks long) JSALT2023 Jelinek Summer Workshop on Speech and Language Technology held in Le Mans this year. Idiap specifically contributed to the theme: Automatic design of conversational models from observation of human-to-human conversation. The final presentations took place on Thu 3rd and Fri 4th August 2023. https://jsalt2023.univ-lemans.fr/en/jsalt-workshop-programme.html contains all links to live streaming.
July 2023: new 3 years long Horizon project ELOQUENCE has been accepted by EC. The call details: HORIZON-CL4-2023-HUMAN-01-03: Natural Language Understanding and Interaction in Advanced Language Technologies (AI Data and Robotics Partnership) (RIA), coordinated by Telefonica, Spain.
July 2023: Two papers accepted for the MDPI Aerospace, Special Issue “Automatic Speech Recognition and Understanding in Air Traffic Management”:
Assistant Based Speech Recognition Support for Air Traffic Controllers in a Multiple Remote Tower Environment, by O. Ohneiser, et al, open access
A Virtual Simulation-Pilot Agent for Training of Air Traffic Controllers, by J. Zuluaga, et al., open access
July 2023: invited speaker on the forensic conference IAFPA 2023
presentation on topic of ROXANNE project to support criminal investigations, and application of technology for forensic speaker comparison.
title: Speaker verification and identification for criminal investigations and forensics.
June 2023: Idiap participates with many submissions on the “Special Issue “Automatic Speech Recognition and Understanding in Air Traffic Management” in MDPI aerospace”, organised by MDPI. More details is here.
June 2023: ROXANNE project was successfuly finalised. Consortium is now about to sign the exploitation plan, allowing to commericalise the technology, share it with partners from civil security domain.
ROXANNE consiortium is preparing the final paper on the Autocrime platform, as well as on the ROXSD database.
The full list of papers resulting from the ROXANNE project is hhere.
June 2023: paper presented by DLR on the ATM seminar, Georgia, USA, 2023.
Our paper was selected amond the best papers.
Title: Automatic Speech Recognition and Understanding for Radar Label Maintenance Support Increases Safety and Reduces Air Traffic Controllers’ Workload.
Authors: Hartmut Helmke, Matthias Kleinert, Nils Ahrenhold, Heiko Ehr, Thorsten Mühlhausen, Oliver Ohneiser, Lucas Klamert, Petr Motlicek, Amrutha Prasad, Juan Pablo Zuluaga-Gomez, Jelena Dokic and Ella Pinska Chauvin.
Presentation is here.
Paper is here.
June 2023: kick-off for TRACY (a big-data analyTics from base-stations Registrations And Cdrs e-evidence sYstem) project, funded under DIGITAL topic: DIGITAL-2022-DEPLOY-02-LAW-SECURITY-AI.
Partners: Performance technologies SAE (Greece), COSMOTE KINITES TILEPIKOINONIES MONOPROSOPI AE (Greece), HELLENIC POLICE (Greece), HOCHSCHULE FUR DEN OFFENTLICHEN DIENST IN BAYERN (Germany), TIME.LEX (Belgium), INSPECTORATUL DE POLITIE JUDETEAN GALATI (Romania), KENTRO MELETON ASFALEIAS (Greece), INSPECTORATUL GENERAL AL POLITIEI (Moldova), Idiap
May 2023: following papers were presented at ICASSP 2023, Greece.
Effectiveness of Text, Acoustic, and Lattice-based representations in Spoken Language Understanding tasks, by E. Villatoro and others.
paper is here
April 2023: Following papers were accepted at Interspeech 2023:
Implementing contextual biasing in GPU decoder for online ASR, by I. Nigmatulina, et al, paper is here.
CommonAccent: Exploring Large Acoustic Pretrained Models for Accent Classification Based on Common Voice, by J. Zuluaga, et al., paper is here.
HyperConformer: Multi-head HyperMixer for Efficient Speech Recognition, by F. Mai, et. al., paper is here.
April 2023: invited presentation on the ASFIC summit organised by Police Judiciaire, Portugal on ROXANNE project:
title: ROXANNE - IA- PLATAFORMA EM TEMPO REAL DE ANÁLISE DE VOZ E TEXTO PARA COMBATER O CRIME ORGANIZADO.
March 2023: ROXANNE (EC H2020 Security project coordinated by Idiap) final review meeting has been organised this month. More info is available at the project website. -March 2023: invited presentation by Armasuisse for the workshop Jahresreporting Forschungsprogramm 2 Kommunikation 2023 on the work of automatic speech recognition and speaker clustering done on the VHF data from ATC domain
title: Speech analytics and language understanding for radio communications
February 2023: Final newsletter of the ROXANNE project has been issued here
February 2023: Official nomination by rector of BUT for the position of associate professor at Faculty of Information Technologies
January 2023: Project consortium meeting of CRiTERIA H2020 EC was organised at Idiap, followed by common team building activities in Swiss mountains
January 2023: New project TRACY: a big-data analyTics from base-stations Registrations And Cdrs e-evidence sYstem, funded by European call DIGITAL-2022-DEPLOY-02 will start in June 2023. Project is coordinated by Performance Technologies
January 2023: New project with EUROCONTROL and Skysoft has started, on integrating speech and text processing technologies to ATC simulator, replacing pseudo-pilots by virtual ones
January 2023: [SLT](https://hk.aconf.org/conf_181262 papers presented:
title: BERTraffic: BERT-based Joint Speaker Role and Speaker Change Detection for Air Traffic Control Communications, authors: Juan Zuluaga-Gomez Seyyed Saeed Sarfjoo, Amrutha Prasad, Iuliia Nigmatulina, Petr Motlicek, Karel Ondrej, Oliver Ohneiser, Hartmut Helmke. paper is here.
title: How Does Pre-trained Wav2Vec 2.0 Perform on Domain-Shifted ASR? An Extensive Benchmark on Air Traffic Control Communications, authors: Zuluaga-Gomez. Juan, Amrutha Prasad, Nigmatulina Iuliia, Seyyed Saeed Sarfjoo, Petr Motlicek, Matthias Kleinert, Hartmut Helmke, Oliver Ohneiser and Qingran Zhan, paper is here.
December 2022: ROXANNE project reaches its end. More information about the final review of the project can be found on Project Website
November 2022: 3 papers were accepted related to our work in HAAWAII project at the SESAR Innovation Days 2022, a conference organised by SJU in Budapest 5-8 Dec, 2022. List of papers:
Speech and Natural Language Processing Technologies for Pseudo-Pilot Simulator, by A. Prasad, et al.
Apron Controller Support by Integration of Automatic Speech Recognition with an Advanced Surface Movement Guidance and Control System, by M. Kleninert, et al.
Grammar Based Speaker Role Identification for Air Traffic Control Speech Recognition, by A. Prasad, et al.
November 2022: HAAWAII EC Horizon project under SJU funding has sufficiently finished. Link to video demonstration is here.
November 2022: Habilitation procedure at BUT, CZ: Public lecture is available here. In-person (room E105, FIT BUT, Božetěchova 2, Brno, CZ) and remote lecture (teams) is planned on November 10, 2022, at 14h CET, link to Teams is here. More info on the website of BUT FIT link.
November 2022: ROXANNE Final Conference in Paris, registration is open here
October 2022: ROXANNE 3rd and Final Field Test conducted successfully!
More details can be found here
Jul 2022: Nomination as a member of the CERIS Expert Group, the Community for European Research and Innovation for Security.
April 2022: Voice technologies for law enforcement and their applications in forensics, lecture at University of Bern, announcement is here
August 2021: preparation of the poster for 30th anniversary of Idiap. The poster is on topic of Ethics in machine learning. Poster is here.
February 2021: Talk given on automatic speech recognition challenges at Swisscom seminar
October 2020: Lecture given at FDP seminar on automatic speech recognition (KIIT School Of Computer Applications, India, Faculty Development Programme (FDP))
June 2020 HAAWAII project kick-off