Ina Kodrasi

I am leading the Signal Processing for Communication Group at Idiap Research Institute, which focuses on establishing signal processing and machine learning approaches to tackle a wide range of challenges arising in human communication.

From the speaker perspective, challenges arise in the presence of speaking impairments. From the listener perspective, challenges arise in the presence of hearing impairments. From the acoustic environment perspective, challenges arise in the presence of undesired interferences such as noise and reverberation.

With these challenges in mind, the high-level objectives of the group are to establish novel digital signal processing and machine learning algorithms for speech, audio, and multimodal signals to automatically detect speaking and hearing impairments, to provide speaking and hearing assistance, and to improve the communication experience in the presence of undesired interferences.

Education & Experience

Publications

Journal Articles

• Cécile Fougeron, Ina Kodrasi, Marina Laganaro, "Differentiation of motor speech disorders through the seven deviance scores from MonPaGe-2.0.s," in Brain Sciences, vol. 12, no. 11, pp. 1471-1487, Oct. 2022.
• Michaela Pernon, Frédéric Assal, Ina Kodrasi, Marina Laganaro, "Perceptual classification of motor speech disorders: the role of severity, speech task, and listener's expertise," in Journal of Speech, Language, and Hearing Research, vol. 65, no. 8, pp. 2727-2747, Aug. 2022.
• Ina Kodrasi, "Temporal envelope and fine structure cues for dysarthric speech detection using CNNs," in IEEE Signal Processing Letters, vol. 28, pp. 1853-1857, Aug. 2021.
• Parvaneh Janbakhshi, Ina Kodrasi, Hervé Bourlard, "Subspace-based learning for automatic dysarthric speech detection," in IEEE Signal Processing Letters, vol. 28, pp. 96-100, Dec. 2020.
• Ina Kodrasi, Hervé Bourlard, "Spectro-temporal sparsity characterization for dysarthric speech detection," in IEEE/ACM Transactions on Audio, Speech and Language Processing, vol. 28, pp. 1210-1222, May 2020.
• Parvaneh Janbakhshi, Ina Kodrasi, Hervé Bourlard, "Automatic pathological speech intelligibility assessment exploiting subspace-based analysis," in IEEE/ACM Transactions on Audio, Speech and Language Processing, vol. 28, pp. 1717-1728, May 2020.
• Ina Kodrasi, Simon Doclo, "Analysis of eigenvalue decomposition-based late reverberation power spectral density estimation," in IEEE/ACM Transactions on Audio, Speech and Language Processing, vol. 26, no. 6, pp. 1106-1118, June 2018. *Best Paper Award, VDE Information Technology Society*
• Ina Kodrasi, Simon Doclo, "Improving the conditioning of the optimization criterion in acoustic multi-channel equalization using shorter reshaping filters," in EURASIP Journal on Advances in Signal Processing, vol. 2018, no. 1, Feb. 2018.
• Ina Kodrasi, Simon Doclo, "Signal-dependent penalty functions for robust acoustic multi-channel equalization," in IEEE/ACM Transactions on Audio, Speech and Language Processing, vol. 25, no. 7, pp. 1512-1525, July 2017.
• Ina Kodrasi, Benjamin Cauchi, Stefan Goetze, Simon Doclo, "Instrumental and perceptual evaluation of dereverberation techniques based on robust acoustic multi-channel equalization," in Journal of the Audio Engineering Society, vol. 65, no. 1/2, pp. 117-129, Jan. 2017.
• Ina Kodrasi, Simon Doclo, "Joint dereverberation and noise reduction based on acoustic multi-channel equalization," in IEEE/ACM Transactions on Audio, Speech and Language Processing, vol. 25, no. 4, pp. 680-693, Apr. 2016.
• Benjamin Cauchi, Ina Kodrasi, Robert Rehr, Stephan Gerlach, Ante Jukic, Timo Gerkmann, Simon Doclo, Stefan Goetze, "Combination of MVDR beamforming and single-channel spectral processing for enhancing noisy and reverberant speech," in EURASIP Journal on Advances in Signal Processing, vol. 2015, no. 1, Dec. 2015.
• Ina Kodrasi, Stefan Goetze, Simon Doclo, "Regularization for partial multichannel equalization for speech dereverberation," in IEEE/ACM Transactions on Audio, Speech and Language Processing, vol. 21, no. 9, pp. 1879-1890, Sept. 2013.

• Mahdi Amiri, Ina Kodrasi, "Suppressing noise disparity in training data for automatic pathological speech detection," in Proc. International Workshop on Acoustic Signal Enhancement, Aalborg, Denmark, September 2024.
• Mahdi Amiri, Ina Kodrasi, "Adversarial Robustness analysis in automatic pathological speech detection approaches," in Proc. Annual Conference of the International Speech Communication Association, Kos Island, Greece, September 2024.
• Mahdi Amiri, Ina Kodrasi, "Test-time adaptation for automatic pathological speech detection in noisy environments," in Proc. European Signal Processing Conference, Lyon, France, August 2024.
• Shakeel Sheikh, Ina Kodrasi, "Impact of speech mode in automatic pathological speech detection," in Proc. European Signal Processing Conference, Lyon, France, August 2024.
• Guilherme Schu, Parvaneh Janbakhshi, Ina Kodrasi, "On using the UA-Speech and TORGO databases to validate automatic dysarthric speech classification approaches ," submitted to Proc. International Conference on Acoustics, Speech, and Signal Processing, Rhodes Island, Greece, June 2023.
• Parvaneh Janbakhshi, Ina Kodrasi, "Adversarial-free speaker identity-invariant representation learning for automatic dysarthric speech classification," in Proc. Annual Conference of the International Speech Communication Association, Incheon, Korea, Sept. 2022.
• Cécile Fougeron, Nicolas Audibert, Ina Kodrasi, Parvaneh Janbakhshi, Michaela Pernon, Nathalie Leveque, Stéphanie Borel, Marina Laganaro, Hérve Bourlard, Frédéric Assal, "Comparison of 5 methods for the evaluation of intelligibility in mild to moderate French dysarthric speech," in Proc. Annual Conference of the International Speech Communication Association, Incheon, Korea, Sept. 2022.
• Parvaneh Janbakhshi, Ina Kodrasi, "Experimental investigation on STFT phase representations for deep learning-based dysarthric speech detection," submitted to IEEE International Conference on Acoustics, Speech, and Signal Processing, Singapore, May 2022, pp. 6477-6481.
• Parvaneh Janbakhshi, Ina Kodrasi, "Supervised speech representation learning for Parkinson's disease classification," in Proc. ITG Conference on Speech Communication, Kiel, Germany, September 2021, pp. 154-158. *Best Paper Award Candidate*
• Lei Wang, Jie Zhu, Ina Kodrasi, "Multi-task single channel speech enhancement using speech presence probability as a secondary task training target," in Proc. European Signal Processing Conference, Dublin, Ireland, August 2021.
• Parvaneh Janbakhshi, Ina Kodrasi, Hervé Bourlard, "Automatic dysarthric speech detection exploiting pairwise distance-based convolutional neural networks," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Toronto, Canada, June 2021, pp. 7328-7332.
• Ina Kodrasi, Michaela Pernon, Marina Laganaro, Hervé Bourlard, "Automatic and perceptual discrimination between dysarthria, apraxia of speech, and neurotypical speech," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Toronto, Canada, June 2021, pp. 7308-7312.
• Ina Kodrasi, Michaela Pernon, Marina Laganaro, Hervé Bourlard, "Automatic discrimination of apraxia of speech and dysarthria using a minimalistic feature set," in Proc. Annual Conference of the International Speech Communication Association, Shanghai, China, October 2020, pp. 4991-4995.
• Parvaneh Janbakhshi, Ina Kodrasi, Hervé Bourlard, "Synthetic speech references for automatic pathological speech intelligibility assessment," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Barcelona, Spain, May 2020, pp. 6099-6103.
• Parvaneh Janbakhshi, Ina Kodrasi, Hervé Bourlard, "Spectral subspace analysis for automatic assessment of pathological speech intelligibility," in Proc. Annual Conference of the International Speech Communication Association, Graz, Austria, September 2019, pp. 3038-3042.
• Marvin Tammen, Ina Kodrasi, Simon Doclo, "Joint estimation of RETF vector and power spectral densities for speech enhancement based on alternating least squares," in Proc. Annual Conference of the International Speech Communication Association, Brighton, United Kingdom, May 2019, pp. 795-799.
• Parvaneh Janbakhshi, Ina Kodrasi, Hervé Bourlard, "Pathological speech intelligibility assessment based on the short-time objective intelligibility measure," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Brighton, United Kingdom, May 2019, pp. 6405-6409.
• Ina Kodrasi, Hervé Bourlard, "Super-Gaussianity of speech spectral coefficients as a potential biomarker for dysarthric speech detection," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Brighton, United Kingdom, May 2019, pp. 6400-6404.
• Ina Kodrasi, Hervé Bourlard, "Statistical modeling of speech spectral coefficients in patients with Parkinson’s disease," in Proc. ITG Conference on Speech Communication, Oldenburg, Germany, October 2018, pp. 271-275.
• Marvin Tammen, Ina Kodrasi, Simon Doclo, "Iterative alternating least-squares approach to jointly estimate the RETFs and the diffuse PSD," in Proc. ITG Conference on Speech Communication, Oldenburg, Germany, October 2018, pp. 221-225.
• Ina Kodrasi, Hervé Bourlard, "Single-channel late reverberation power spectral density estimation using denoising autoencoders," in Proc. Annual Conference of the International Speech Communication Association, Hyderabad, India, September 2018, pp. 1319-1323.
• Gilles Courtois, Vincent Grimaldi, Hervé Lissek, Ina Kodrasi, Eleftheria Georganti, "Experimental evaluation of speech enhancement methods in remote microphone systems for hearing aids," in Proc. European Congress and Exposition on Noise Control Engineering, Crete, Greece, May 2018, pp. 351-358.
• Ina Kodrasi, Simon Doclo, "Joint late reverberation and noise power spectral density estimation in a spatially homogeneous noise field," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Calgary, Canada, March 2018, pp. 441-445.
• Marvin Tammen, Ina Kodrasi, Simon Doclo, "Complexity reduction of eigenvalue decomposition-based diffuse power spectral density estimators using the power method," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Calgary, Canada, March 2018, pp. 451-455.
• Ina Kodrasi, Simon Doclo, "Multi-channel late reverberation power spectral density estimation based on nuclear norm minimization," in Proc. IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, New Paltz, USA, October 2017, pp. 101-105.
• Ina Kodrasi, Daniel Marquardt, Simon Doclo, "A simulation study on binaural dereverberation and noise reduction based on diffuse power spectral density estimators," in Proc. International Workshop on Challenges in Hearing Assistive Technology, Stockholm, Sweden, August 2017.
• Ina Kodrasi, Simon Doclo, "Late reverberant power spectral density estimation based on an eigenvalue decomposition," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, New Orleans, USA, March 2017, pp. 611-615.
• Ina Kodrasi, Simon Doclo, "EVD-based multi-channel dereverberation of a moving speaker using different RETF estimation methods," in Proc. Hands-Free Speech Communication and Microphone Arrays, San Francisco, USA, March 2017, pp. 116-120.
• Ina Kodrasi, Ante Jukic, Simon Doclo, "Robust sparsity-promoting acoustic multi-channel equalization for speech dereverberation," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Shanghai, China, March 2016, pp. 166-170.
• Ina Kodrasi, Simon Doclo, "Incorporating the noise statistics in acoustic multi-channel equalization," in Proc. AES Conference on Dereverberation and Reverberation of Audio, Music, and Speech, Leuven, Belgium, February 2016.
• Ina Kodrasi, Daniel Marquardt, Simon Doclo, "Curvature-based optimization of the trade-off parameter in the speech distortion weighted multichannel Wiener filter," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Brisbane, Australia, April 2015, pp. 315-319.
• Ina Kodrasi, Simon Doclo, "Joint dereverberation and noise reduction based on acoustic multichannel equalization," in Proc. International Workshop on Acoustic Signal Enhancement, Juan les Pins, France, September 2014, pp. 140-144.
• Stefan Goetze, Anna Warzybok, Ina Kodrasi, Jan Ole Jungmann, Benjamin Cauchi, Jan Rennies, Emanuel Habets, Alfred Mertins, Timo Gerkmann, Simon Doclo, Birger Kollmeier, "A study on speech quality and speech intelligibility measures for quality assessment of single-channel dereverberation algorithms," in Proc. International Workshop on Acoustic Signal Enhancement, Juan les Pins, France, September 2014, pp. 234-238.
• Anna Warzybok, Ina Kodrasi, Jan Ole Jungmann, Emanuel Habets, Timo Gerkmann, Alfred Mertins, Simon Doclo, Birger Kollmeier, Stefan Goetze, "Subjective speech quality and speech intelligibility evaluation of single-channel dereverberation algorithms," in Proc. International Workshop on Acoustic Signal Enhancement, Juan les Pins, France, September 2014, pp. 333-337.
• Benjamin Cuachi, Ina Kodrasi, Robert Rehr, Stephan Gerlach, Ante Jukic, Timo Gerkmann, Simon Doclo, Stefan Goetze, "Joint dereverberation and noise reduction using beamforming and a single-channel speech enhancement scheme," in Proc. IEEE REverberant Voice Enhancement and Recognition Benchmark Challenge (REVERB), Florence, Italy, May 2014.
• Ina Kodrasi, Timo Gerkmann, Simon Doclo, "Frequency-domain single-channel inverse filtering for speech dereverberation: Theory and practice," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Florence, Italy, May 2014, pp. 5214-5218.
• Ina Kodrasi, Simon Doclo, "Regularized subspace-based acoustic multichannel equalization for speech dereverberation," in Proc. European Signal Processing Conference, Marrakesh, Morocco, September 2013.
• Ina Kodrasi, Stefan Goetze, Simon Doclo, "A perceptually constrained channel shortening technique for speech dereverberation," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Vancouver, Canada, May 2013, pp. 151-155. *Best Paper Award Candidate*
• Ina Kodrasi, Stefan Goetze, Simon Doclo, "Non-intrusive regularization for least-squares multichannel equalization techniques for speech dereverberation," in Proc. EEE Convention of Electrical and Electronics Engineers in Israel, Eilat, Israel, November 2012.
• Ina Kodrasi, Stefan Goetze, Simon Doclo, "Increasing the robustness of acoustic multichannel equalization by means of regularization," in Proc. International Workshop on Acoustic Signal Enhancement, Aachen, Germany, September 2012, pp. 161-164.
• Ina Kodrasi, Simon Doclo, "The effect of inverse filter length on the robustness of acoustic multichannel equalization," in Proc. European Signal Processing Conference, Bucharest, Romania, August 2012, pp. 2442-2446.
• Ina Kodrasi, Simon Doclo, "Robust partial multichannel equalization techniques for speech dereverberation," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Kyoto, Japan, March 2012, pp. 537-540.
• Ina Kodrasi, Simon Doclo, "Microphone position optimization for planar superdirective beamforming," in Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing, Prague, Czech Republic, May 2011, pp. 109-112.
• Werner Henkel, Fanging Hu, Ina Kodrasi, "Inherent time-frequency coding in OFDM - a Possibility for ISI correction without a cyclic prefix?," in Proc. International OFDM Workshop, Hamburg, Germany, September 2009.
• Tim Fuehner, Christian Campen, Ina Kodrasi, Alexander Burenkov, Andreas Erdmann, "A simulation study on the impact of lithographic process variations on CMOS device performance," in Proc. of SPIE - The International Society for Optical Engineering, California, USA, April 2008.

Books

• Ina Kodrasi, "Dereverberation and noise reduction techniques based on acoustic multi-channel equalization," Dr. Hut, Oldenburg, Germany, 2015. *EURASIP Best PhD Award*

Teaching

Projects

To assist the clinical diagnosis and treatment of motor speech disorders through objective and replicable results, we worked on devising pathological speech processing techniques to (semi-) automatically classify typical and pathological speech, to discriminate between subtypes of disorders, and to assess speech intelligibility, speech quality, and the severity of the disorder. More specifically, we focused on (1) automatic analysis of pathological speech by means of automatic speech recognition (ASR) systems, (2) handcrafting distinct acoustic markers which can robustly capture different impaired speech dimensions, and (3) leveraging data-driven deep learning approaches.

This research was supported by the Swiss National Science Foundation, project no CRSII5_173711, in collaboration with University of Geneva, University of Paris 3-Sorbonne Nouvelle, and Geneva University Hospitals.

Academic Activties