2 Post-doctoral positions at Le Mans University on Deep learning approaches speech processing

*Place of work* Le Mans University, Le Mans – France
*Starting date* From now to June 2020
*Salary* between 2 300 and 2 600 ¿ /month
*Duration* 12 months and 24 months (can be combined in a 36 months position)


**1st position

* Context *
            The LST team from LIUM (Le Mans University) is focusing on autonomous system’s behavior
            for the task of speaker diarization and machine translation.
            The ALLIES project (European Chist-ERA collaborative project) aims at developing
            evaluation protocols, metrics and scenarios for lifelong learning autonomous systems.
            The goal is to enable auto-adaptable systems that can also auto-evaluate in order to
            sustain their performance across time. Autonomous systems can rely on human domain
            experts via active and interactive learning processes to be define within the ALLIES project.

            * Missions *
            Develop an autonomous system for speaker diarization by integrating lifelong learning,
            active and interactive learning components. The research work will be related to some of the following topics:
            - unsupervised adaptation
            - unsupervised evaluation
            - active learning (based on the unsupervised evaluation process, the autonomous
                  system is free to require additional knowledge from the human domain expert)
            - Interactive learning (a human domain expert provides specific knowledge to
                  the autonomous system. This information must be taken into account by the system)
            Performance will be analyzed using protocols, metrics and scenarios developed for the ALLIES project.

            * Duration * 12 months
            * Salary * 2 365,14¿ (after taxes)
            * Start * as soon as possible, latest January 2020
            * Supervisers * Anthony Larcher (anthony.larcher@univ-lemans.fr) and Loïc Barrault (loic.barrault@univ-lemans.fr)

            Expected competences:
                    - Phd in Machine Learning and Deep Learning
                    - Experience in speech processing is positive
                    - Python fluent
                    - familiar with a deep learning toolkit (Pytorch, TensorFlow)

**2nd position
            * Context *
            The LST team from LIUM (Le Mans University) is focusing on evolutive end-to-end
            neural networks for speaker recognition. The Extensor project (French ANR funded)
            aims at developing novel architectures for end-to-end speaker recognition as well as
            explaining the behavior of those networks. The focus of Extensor is threefold:
            get rid of the legacy of bayesian system’s architecture and explore wider opportunities offered in deep learning;
            explore real end-to-end architectures exploiting the tax signal instead of classical features (such as MFCC of filterbanks);
            Develop tools for explainability in speaker recognition.

            * Missions *
            Develop end-to-end speaker recognition system based on state-of-the-art approaches (x-vectors, sincnet…)
            Develop evolutive architectures making use of existing genetic algorithms and study their behavior.
            Participate to the three hackathons organized by the Extensor project in order to develop
            tools for evolutive neural network architecture and explainability for speaker recognition.
            Dissemination: the research will be published in the major conferences and journals

            * Duration * 24 months
            * Salary * 2 600¿ (after taxes)
            * Start * as soon as possible, latest June 2020
            * Location * LIUM, Le Mans University
            * Superviser * Anthony Larcher (anthony.larcher@univ-lemans.fr)

            Expected competences:
                    - Phd in Machine Learning and Deep Learning
                    - Experience in speech processing is positive
                    - Python fluent
                    - familiar with a deep learning toolkit (Pytorch, TensorFlow)