@inproceedings{StemmerGeorgesHoferetal.2017, author = {Stemmer, Georg and Georges, Munir and Hofer, Joachim and Rozen, Piotr and Bauer, Josef and Nowicki, Jakub and Bocklet, Tobias and Colett, Hannah R. and Falik, Ohad and Deisher, Michael and Downing, Sylvia J.}, title = {Speech Recognition and Understanding on Hardware-Accelerated DSP}, booktitle = {Proceedings Interspeech 2017}, publisher = {International Speech Communication Association (ISCA)}, address = {Baixas}, url = {https://www.isca-speech.org/archive/interspeech_2017/stemmer17_interspeech.html}, doi = {https://doi.org/10.21437/Interspeech.2017}, pages = {2036 -- 2037}, year = {2017}, language = {en} } @inproceedings{ChenDeisherGeorges2023, author = {Chen, Liu and Deisher, Michael and Georges, Munir}, title = {An End-to-End Neural Network for Image-to-Audio Transformation}, booktitle = {Proceedings of the 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}, publisher = {IEEE}, address = {Piscataway}, isbn = {978-1-7281-6327-7}, doi = {https://doi.org/10.1109/ICASSP49357.2023.10096121}, year = {2023}, language = {en} } @inproceedings{GeorgesKanthakKlakow2014, author = {Georges, Munir and Kanthak, Stephan and Klakow, Dietrich}, title = {Accurate client-server based speech recognition keeping personal data on the client}, booktitle = {2014 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP)}, publisher = {IEEE}, address = {Piscataway}, isbn = {978-1-4799-2893-4}, issn = {2379-190X}, doi = {https://doi.org/10.1109/ICASSP.2014.6854205}, pages = {3271 -- 3275}, year = {2014}, language = {en} } @inproceedings{GeorgesKanthakKlakow2013, author = {Georges, Munir and Kanthak, Stephan and Klakow, Dietrich}, title = {Transducer-based speech recognition with dynamic language models}, booktitle = {Proceedings Interspeech 2013}, publisher = {International Speech Communication Association (ISCA)}, address = {Baixas}, doi = {doi.org/10.21437/Interspeech.2013-185}, pages = {642 -- 646}, year = {2013}, language = {en} } @inproceedings{HerygersVerkhodanovaColeretal.2023, author = {Herygers, Aaricia and Verkhodanova, Vass and Coler, Matt and Scharenborg, Odette and Georges, Munir}, title = {Bias in Flemish Automatic Speech Recognition}, booktitle = {Elektronische Sprachsignalverarbeitung 2023: Tagungsband der 34. Konferenz M{\"u}nchen, 1.-3. M{\"a}rz 2023}, editor = {Draxler, Christoph}, publisher = {TUDpress}, address = {Dresden}, isbn = {978-3-95908-303-4}, url = {https://www.essv.de/paper.php?id=1186}, pages = {158 -- 165}, year = {2023}, language = {en} } @inproceedings{KendrickFrohnmaierGeorges2021, author = {Kendrick, Caroline and Frohnmaier, Mariano and Georges, Munir}, title = {Audio-Visual Recipe Guidance for Smart Kitchen Devices}, booktitle = {ICNLSP 2021: Proceedings of the 4th International Conference on Natural Language and Speech Processing}, publisher = {ACL}, address = {Stroudsburg}, isbn = {978-1-955917-18-6}, doi = {https://aclanthology.org/2021.icnlsp-1.30}, pages = {257 -- 261}, year = {2021}, language = {en} } @inproceedings{GeorgesCzarnowskiBocklet2019, author = {Georges, Munir and Czarnowski, Krzysztof and Bocklet, Tobias}, title = {Ultra-Compact NLU}, booktitle = {Proceedings Interspeech 2019}, subtitle = {Neuronal Network Binarization as Regularization}, publisher = {ISCA}, address = {Baixas}, doi = {https://doi.org/10.21437/Interspeech.2019-2591}, pages = {809 -- 813}, year = {2019}, language = {en} } @inproceedings{HartungHerygersKurlekaretal.2023, author = {Hartung, Kai and Herygers, Aaricia and Kurlekar, Shubham Vijay and Zakaria, Khabbab and Volkan, Taylan and Gr{\"o}ttrup, S{\"o}ren and Georges, Munir}, title = {Measuring Sentiment Bias in Machine Translation}, booktitle = {Text, Speech, and Dialogue: 26th International Conference: Proceedings}, editor = {Ekštein, Kamil and P{\´a}rtl, František and Konop{\´i}k, Miloslav}, publisher = {Springer}, address = {Cham}, isbn = {978-3-031-40498-6}, issn = {1611-3349}, doi = {https://doi.org/10.1007/978-3-031-40498-6_8}, pages = {82 -- 93}, year = {2023}, language = {en} } @inproceedings{GlockerHerygersGeorges2023, author = {Glocker, Kevin and Herygers, Aaricia and Georges, Munir}, title = {Allophant: Cross-lingual Phoneme Recognition with Articulatory Attributes}, booktitle = {INTERSPEECH 2023}, publisher = {ISCA}, address = {Baixas}, doi = {https://doi.org/10.21437/Interspeech.2023-772}, pages = {2258 -- 2262}, year = {2023}, language = {en} } @inproceedings{SrinivasaganDeisherGeorges2023, author = {Srinivasagan, Gokul and Deisher, Michael and Georges, Munir}, title = {Compression of end-to-end non-autoregressive image-to-speech system for lowresourced devices}, booktitle = {Speech Communication, 15th ITG Conference, 20.-22.09.2023 in Aachen, Germany}, publisher = {VDE Verlag}, address = {Berlin}, isbn = {978-3-8007-6164-7}, doi = {https://doi.org/10.30420/456164029}, pages = {151 -- 155}, year = {2023}, language = {en} } @inproceedings{RanzenbergerBockletFreisingeretal.2023, author = {Ranzenberger, Thomas and Bocklet, Tobias and Freisinger, Steffen and Frischholz, Lia and Georges, Munir and Glocker, Kevin and Herygers, Aaricia and Peinl, Ren{\´e} and Riedhammer, Korbinian and Schneider, Fabian and Simic, Christopher and Zakaria, Khabbab}, title = {The Hochschul-Assistenz-System HAnS: An ML-Based Learning Experience Platform}, booktitle = {Elektronische Sprachsignalverarbeitung 2023: Tagungsband der 34. Konferenz M{\"u}nchen, 1.-3. M{\"a}rz 2023}, editor = {Draxler, Christoph}, publisher = {TUDpress}, address = {Dresden}, isbn = {978-3-95908-303-4}, url = {https://www.essv.de/paper.php?id=1188}, pages = {168 -- 169}, year = {2023}, language = {en} } @inproceedings{FreisingerSchneiderHerygersetal.2023, author = {Freisinger, Steffen and Schneider, Fabian and Herygers, Aaricia and Georges, Munir and Bocklet, Tobias and Riedhammer, Korbinian}, title = {Unsupervised Multilingual Topic Segmentation of Video Lectures: What can Hierarchical Labels tell us about the Performance?}, booktitle = {Proceedings 9th Workshop on Speech and Language Technology in Education (SLaTE)}, publisher = {International Speech Communication Association (ISCA)}, address = {Baixas}, doi = {https://doi.org/10.21437/SLaTE.2023-27}, pages = {141 -- 145}, year = {2023}, language = {en} } @inproceedings{HartungJaegerGroettrupetal.2022, author = {Hartung, Kai and J{\"a}ger, Gerhard and Gr{\"o}ttrup, S{\"o}ren and Georges, Munir}, title = {Typological Word Order Correlations with Logistic Brownian Motion}, pages = {2022.sigtyp-1.3}, booktitle = {Proceedings of the 4th Workshop on Computational Typology and Multilingual NLP (SIGTYP 2022)}, editor = {Vylomova, Ekaterina and Ponti, Edoardo and Cotterell, Ryan}, publisher = {Association for Computational Linguistics}, address = {Stroudsburg}, isbn = {978-1-955917-93-3}, url = {https://aclanthology.org/2022.sigtyp-1.3}, pages = {22 -- 26}, year = {2022}, abstract = {In this study we address the question to what extent syntactic word-order traits of different languages have evolved under correlation and whether such dependencies can be found universally across all languages or restricted to specific language families.To do so, we use logistic Brownian Motion under a Bayesian framework to model the trait evolution for 768 languages from 34 language families. We test for trait correlations both in single families and universally over all families. Separate models reveal no universal correlation patterns and Bayes Factor analysis of models over all covered families also strongly indicate lineage specific correlation patters instead of universal dependencies.}, language = {en} } @unpublished{GlockerHerygersGeorges2023, author = {Glocker, Kevin and Herygers, Aaricia and Georges, Munir}, title = {Allophant: Cross-lingual Phoneme Recognition with Articulatory Attributes}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2306.04306}, year = {2023}, abstract = {This paper proposes Allophant, a multilingual phoneme recognizer. It requires only a phoneme inventory for crosslingual transfer to a target language, allowing for low-resource recognition. The architecture combines a compositional phone embedding approach with individually supervised phonetic attribute classifiers in a multi-task architecture. We also introduce Allophoible, an extension of the PHOIBLE database. When combined with a distance based mapping approach for grapheme-to-phoneme outputs, it allows us to train on PHOIBLE inventories directly. By training and evaluating on 34 languages, we found that the addition of multi-task learning improves the model's capability of being applied to unseen phonemes and phoneme inventories. On supervised languages we achieve phoneme error rate improvements of 11 percentage points (pp.) compared to a baseline without multi-task learning. Evaluation of zero-shot transfer on 84 languages yielded a decrease in PER of 2.63 pp. over the baseline.}, language = {en} } @inproceedings{GeorgesHuangBocklet2020, author = {Georges, Munir and Huang, Jonathan and Bocklet, Tobias}, title = {Compact Speaker Embedding: lrx-vector}, booktitle = {Proceedings Interspeech 2020}, publisher = {ISCA}, address = {[s. l.]}, doi = {https://doi.org/10.21437/Interspeech.2020-2106}, pages = {3236 -- 3240}, year = {2020}, language = {en} } @unpublished{PagonisHartungWuetal.2024, author = {Pagonis, Panagiotis and Hartung, Kai and Wu, Di and Georges, Munir and Gr{\"o}ttrup, S{\"o}ren}, title = {Analysis of Knowledge Tracing performance on synthesised student data}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2401.16832}, year = {2024}, abstract = {Knowledge Tracing (KT) aims to predict the future performance of students by tracking the development of their knowledge states. Despite all the recent progress made in this field, the application of KT models in education systems is still restricted from the data perspectives: 1) limited access to real life data due to data protection concerns, 2) lack of diversity in public datasets, 3) noises in benchmark datasets such as duplicate records. To resolve these problems, we simulated student data with three statistical strategies based on public datasets and tested their performance on two KT baselines. While we observe only minor performance improvement with additional synthetic data, our work shows that using only synthetic data for training can lead to similar performance as real data.}, language = {en} } @inproceedings{PagonisHartungWuetal.2024, author = {Pagonis, Panagiotis and Hartung, Kai and Wu, Di and Georges, Munir and Gr{\"o}ttrup, S{\"o}ren}, title = {Analysis of Knowledge Tracing performance on synthesised student data}, publisher = {Universit{\"a}t Bamberg}, address = {Bamberg}, url = {https://sme.uni-bamberg.de/ai4ai/}, year = {2024}, language = {en} } @inproceedings{GlockerGeorges2022, author = {Glocker, Kevin and Georges, Munir}, title = {Hierarchical Multi-Task Transformers for Crosslingual Low Resource Phoneme Recognition}, booktitle = {Proceedings of the 5th International Conference on Natural Language and Speech Processing (ICNLSP 2022)}, publisher = {Association for Computational Linguistics}, address = {Stroudsburg}, isbn = {978-1-959429-36-4}, url = {https://aclanthology.org/2022.icnlsp-1.21}, pages = {187 -- 192}, year = {2022}, abstract = {This paper proposes a method for multilingual phoneme recognition in unseen, low resource languages. We propose a novel hierarchical multi-task classifier built on a hybrid convolution-transformer acoustic architecture where articulatory attribute and phoneme classifiers are optimized jointly. The model was evaluated on a subset of 24 languages from the Mozilla Common Voice corpus. We found that when using regular multi-task learning, negative transfer effects occurred between attribute and phoneme classifiers. They were reduced by the hierarchical architecture. When evaluating zero-shot crosslingual transfer on a data set with 95 languages, our hierarchical multi-task classifier achieves an absolute PER improvement of 2.78\% compared to a phoneme-only baseline.}, language = {en} } @inbook{GlockerGeorges2023, author = {Glocker, Kevin and Georges, Munir}, title = {Hierarchical Multi-task Learning with Articulatory Attributes for Cross-Lingual Phoneme Recognition}, booktitle = {Practical Solutions for Diverse Real-World NLP Applications}, editor = {Abbas, Mourad}, publisher = {Springer}, address = {Cham}, isbn = {978-3-031-44259-9}, doi = {https://doi.org/10.1007/978-3-031-44260-5_4}, pages = {59 -- 75}, year = {2023}, language = {en} } @inproceedings{RanzenbergerBockletFreisingeretal.2024, author = {Ranzenberger, Thomas and Bocklet, Tobias and Freisinger, Steffen and Georges, Munir and Glocker, Kevin and Herygers, Aaricia and Riedhammer, Korbinian and Schneider, Fabian and Simic, Christopher and Zakaria, Khabbab}, title = {Extending HAnS: Large Language Models for Question Answering, Summarization, and Topic Segmentation in an ML-based Learning Experience Platform}, booktitle = {Elektronische Sprachsignalverarbeitung 2024, Tagungsband der 35. Konferenz, Regensburg, 6.-8. M{\"a}rz 2024}, editor = {Baumann, Timo}, publisher = {TUDPress}, address = {Dresden}, isbn = {978-3-95908-325-6}, doi = {https://doi.org/10.35096/othr/pub-7103}, pages = {219 -- 224}, year = {2024}, language = {en} }