Detecting Infant Cries in Noisy NICU Recordings: A Study on SSL Models.

Péter Mihajlik https://orcid.org/0000-0001-7532-9773 Ákos Antal Péter Földesy Péter Mihajlik Detecting Infant Cries in Noisy NICU Recordings: A Study on SSL Models. 151-159 2026 BIOSTEC (1) https://doi.org/10.5220/0014458000004070 conf/biostec/2026-1 db/conf/biostec/biostec2026-1.html#AntalFM26

Máté Gedeon Péter Mihajlik Speaker-Aware Simulation Improves Conversational Speech Recognition. 2026 February abs/2602.04776 CoRR https://doi.org/10.48550/arXiv.2602.04776 db/journals/corr/corr2602.html#abs-2602-04776 streams/journals/corr

Máté Gedeon Piroska Barta Péter Mihajlik Katalin Mády Scaling Conversational Hungarian ASR: The BEA-Dialogue+ Corpus. 2026 May abs/2605.31469 CoRR https://doi.org/10.48550/arXiv.2605.31469 db/journals/corr/corr2605.html#abs-2605-31469 streams/journals/corr

Máté Gedeon Péter Mihajlik Efficient ASR Training with Conversations that Never Happened. 2026 June abs/2606.03957 CoRR https://doi.org/10.48550/arXiv.2606.03957 db/journals/corr/corr2606.html#abs-2606-03957 streams/journals/corr

Péter Mihajlik Éva Székely Piroska Barta Máté Soma Kádár Gergely Dobsinszki László Tóth 0001 Improved Dysarthric Speech to Text Conversion via TTS Personalization. 521-525 2025 EUSIPCO https://ieeexplore.ieee.org/document/11226704 conf/eusipco/2025 db/conf/eusipco/eusipco2025.html#MihajlikSBKDT25 streams/conf/eusipco Éva Székely Péter Mihajlik Máté Soma Kádár László Tóth 0001 Voice Reconstruction through Large-Scale TTS Models: Comparing Zero-Shot and Fine-tuning Approaches to Personalise TTS in Assistive Communication. 2025 INTERSPEECH https://doi.org/10.21437/Interspeech.2025-1726 conf/interspeech/2025 db/conf/interspeech/interspeech2025.html#SzekelyMK025 Yue Luo Xihang Xia Kinga Jelencsik-Mátyus Shaimaa Alwaisi Péter Mihajlik From CHAT towards ASR: A Hybrid Pipeline for Constructing the HUKILC-CO Hungarian Child Speech Dataset. 1-5 2025 IWSSIP https://doi.org/10.1109/IWSSIP66997.2025.11151863 conf/iwssip/2025 db/conf/iwssip/iwssip2025.html#LuoXJAM25 streams/conf/iwssip Gergely Dobsinszki Péter Mihajlik Mate S. Kadar Tibor Fegyó Katalin Mády Best Data is more Supervised Data - Even for Hungarian ASR. 60-69 2025 SPECOM (2) https://doi.org/10.1007/978-3-032-07959-6_5 conf/specom/2025-2 db/conf/specom/specom2025-2.html#DobsinszkiMKFM25 streams/conf/specom Piroska Barta Péter Mihajlik Towards Hungarian-English Code-Switching Speech Dataset Construction by using Multi Speaker-adaptive Text-to-Speech Synthesis. 81-85 2025 SpeD https://doi.org/10.1109/SpeD67700.2025.11251954 conf/sped/2025 db/conf/sped/sped2025.html#BartaM25 streams/conf/sped Dalai Mengke Wuyerigen Péter Mihajlik Impact of Text Origin and Real-Synthetic Data Ratio in TTS-Augmented Low-Resource ASR. 97-101 2025 SpeD https://doi.org/10.1109/SpeD67700.2025.11252042 conf/sped/2025 db/conf/sped/sped2025.html#MengkeWM25 streams/conf/sped Amin Hassairi Péter Mihajlik Cross-lingual Transfer Learning Experiments for Arabic ASR. 108-113 2025 SpeD https://doi.org/10.1109/SpeD67700.2025.11252407 conf/sped/2025 db/conf/sped/sped2025.html#HassairiM25 streams/conf/sped Dalai Mengke Yan Meng Péter Mihajlik How Far Can Synthetic Speech Go? Enhancing ASR in Low-Resource Scenarios via Voice Cloning. 207-217 2025 TSD (1) https://doi.org/10.1007/978-3-032-02548-7_18 conf/tsd/2025-1 db/conf/tsd/tsd2025-1.html#MengkeMM25 streams/conf/tsd Yan Meng Péter Mihajlik Dalai Mengke Katalin Mády Robust Disfluency Labeling in Spontaneous Speech: Insights from Diverse Hungarian Corpora Including Atypical Speakers. 289-298 2025 TSD (1) https://doi.org/10.1007/978-3-032-02548-7_24 conf/tsd/2025-1 db/conf/tsd/tsd2025-1.html#MengMMM25 streams/conf/tsd

Péter Mihajlik Éva Székely Piroska Barta Mate S. Kadar Gergely Dobsinszki László Tóth 0001 Improved Dysarthric Speech to Text Conversion via TTS Personalization. 2025 August abs/2508.06391 CoRR https://doi.org/10.48550/arXiv.2508.06391 db/journals/corr/corr2508.html#abs-2508-06391 streams/journals/corr

Máté Gedeon Péter Mihajlik From Independence to Interaction: Speaker-Aware Simulation of Multi-Speaker Conversational Timing. 2025 September abs/2509.15808 CoRR https://doi.org/10.48550/arXiv.2509.15808 db/journals/corr/corr2509.html#abs-2509-15808 streams/journals/corr

Máté Gedeon Péter Mihajlik LibriConvo: Simulating Conversations from Read Literature for ASR and Diarization. 2025 October abs/2510.23320 CoRR https://doi.org/10.48550/arXiv.2510.23320 db/journals/corr/corr2510.html#abs-2510-23320 streams/journals/corr

Máté Gedeon Piroska Barta Péter Mihajlik Tekla Etelka Gráczi Anna Kohári Katalin Mády Toward Conversational Hungarian Speech Recognition: Introducing the BEA-Large and BEA-Dialogue Datasets. 2025 November abs/2511.13529 CoRR https://doi.org/10.48550/arXiv.2511.13529 db/journals/corr/corr2511.html#abs-2511-13529 streams/journals/corr

Péter Mihajlik Katalin Mády Anna Kohári Fruzsina Sára Fruzsina Gábor Kiss Tekla Etelka Gráczi A. Seza Dogruöz Is Spoken Hungarian Low-resource?: A Quantitative Survey of Hungarian Speech Data Sets. 9382-9388 2024 LREC/COLING https://aclanthology.org/2024.lrec-main.820 conf/coling/2024 db/conf/coling/coling2024.html#MihajlikMKFKGD24 Péter Mihajlik Yan Meng Mate S. Kadar Julian Linke Barbara Schuppler Katalin Mády On Disfluency and Non-lexical Sound Labeling for End-to-end Automatic Speech Recognition. 2024 INTERSPEECH https://doi.org/10.21437/Interspeech.2024-2157 conf/interspeech/2024 db/conf/interspeech/interspeech2024.html#MihajlikMKLSM24 Yue Luo Péter Mihajlik Comparison of Well and Lower-Resourced Self-training in ASR. 45-56 2024 SPECOM (1) https://doi.org/10.1007/978-3-031-77961-9_3 conf/specom/2024-1 db/conf/specom/specom2024-1.html#LuoM24 streams/conf/specom Julian Linke Mate Kadar Gergely Dosinszky Péter Mihajlik Gernot Kubin Barbara Schuppler What do self-supervised speech representations encode? An analysis of languages, varieties, speaking styles and speakers. 5371-5375 2023 INTERSPEECH https://doi.org/10.21437/Interspeech.2023-951 conf/interspeech/2023 db/conf/interspeech/interspeech2023.html#LinkeKDMKS23 Péter Mihajlik András Balog Tekla Etelka Gráczi Anna Kohári Balázs Tarján Katalin Mády BEA-Base: A Benchmark for ASR of Spontaneous Hungarian. 1970-1977 2022 LREC https://aclanthology.org/2022.lrec-1.211 conf/lrec/2022 db/conf/lrec/lrec2022.html#MihajlikBGKTM22

Péter Mihajlik András Balog Tekla Etelka Gráczi Anna Kohári Balázs Tarján Katalin Mády BEA-Base: A Benchmark for ASR of Spontaneous Hungarian. 2022 abs/2202.00601 CoRR https://arxiv.org/abs/2202.00601 db/journals/corr/corr2202.html#abs-2202-00601

Balázs Tarján György Szaszák Tibor Fegyó Péter Mihajlik Improving Real-time Recognition of Morphologically Rich Speech with Transformer Language Model. 491-496 2020 CogInfoCom https://doi.org/10.1109/CogInfoCom50765.2020.9237817 conf/coginfocom/2020 db/conf/coginfocom/coginfocom2020.html#TarjanSFM20 Balázs Tarján György Szaszák Tibor Fegyó Péter Mihajlik On the Effectiveness of Neural Text Generation Based Data Augmentation for Recognition of Morphologically Rich Speech. 437-445 2020 TDS https://doi.org/10.1007/978-3-030-58323-1_47 conf/tsd/2020 db/conf/tsd/tsd2020.html#TarjanSFM20

Balázs Tarján György Szaszák Tibor Fegyó Péter Mihajlik On the Effectiveness of Neural Text Generation based Data Augmentation for Recognition of Morphologically Rich Speech. 2020 abs/2006.05129 CoRR https://arxiv.org/abs/2006.05129 db/journals/corr/corr2006.html#abs-2006-05129

Balázs Tarján György Szaszák Tibor Fegyó Péter Mihajlik Deep Transformer based Data Augmentation with Subword Units for Morphologically Rich Online ASR. 2020 abs/2007.06949 CoRR https://arxiv.org/abs/2007.06949 db/journals/corr/corr2007.html#abs-2007-06949

Balázs Tarján György Szaszák Tibor Fegyó Péter Mihajlik N-gram Approximation of LSTM Recurrent Language Models for Single-pass Recognition of Hungarian Call Center Conversations. 131-136 2019 CogInfoCom https://doi.org/10.1109/CogInfoCom47531.2019.9089959 conf/coginfocom/2019 db/conf/coginfocom/coginfocom2019.html#TarjanSFM19 Balázs Tarján György Szaszák Tibor Fegyó Péter Mihajlik Investigation on N-Gram Approximated RNNLMs for Recognition of Morphologically Rich Speech. 223-234 2019 SLSP https://doi.org/10.1007/978-3-030-31372-2_19 conf/slsp/2019 db/conf/slsp/slsp2019.html#TarjanSFM19

Balázs Tarján György Szaszák Tibor Fegyó Péter Mihajlik Investigation on N-gram Approximated RNNLMs for Recognition of Morphologically Rich Speech. 2019 abs/1907.06407 CoRR http://arxiv.org/abs/1907.06407 db/journals/corr/corr1907.html#abs-1907-06407

Péter Mihajlik Lili Szabó Balázs Tarján András Balog Krisztina Rábai First Results in Developing a Medieval Latin Language Charter Dictation System for the East-Central Europe Region. 2058-2062 2017 INTERSPEECH https://doi.org/10.21437/Interspeech.2017-1558 conf/interspeech/2017 db/conf/interspeech/interspeech2017.html#MihajlikSTBR17 Lili Szabó Péter Mihajlik András Balog Tibor Fegyó Unified Simplified Grapheme Acoustic Modeling for Medieval Latin LVCSR. 420-428 2017 TSD https://doi.org/10.1007/978-3-319-64206-2_47 conf/tsd/2017 db/conf/tsd/tsd2017.html#SzaboMBF17 Ádam Varga Balázs Tarján Zoltán Tobler György Szaszák Tibor Fegyó Csaba Bordás Péter Mihajlik Automatic Close Captioning for Live Hungarian Television Broadcast Speech: A Fast and Resource-Efficient Approach. 105-112 2015 SPECOM https://doi.org/10.1007/978-3-319-23132-7_13 conf/specom/2015 db/conf/specom/specom2015.html#VargaTTSFBM15

Gellért Sárosi Balázs Tarján Tibor Fegyó Péter Mihajlik Automated transcription of conversational Call Center speech - with respect to non-verbal acoustic events. 265-275 2014 8 Intell. Decis. Technol. 4 https://doi.org/10.3233/IDT-140195 db/journals/idt/idt8.html#SarosiTFM14

Lili Szabó Balázs Tarján Péter Mihajlik Writing with speech: A qualitative user evaluation study. 355-360 2014 CogInfoCom https://doi.org/10.1109/CogInfoCom.2014.7020478 conf/coginfocom/2014 db/conf/coginfocom/coginfocom2014.html#SzaboTM14 Balázs Tarján Tibor Fegyó Péter Mihajlik A bilingual study on the prediction of morph-based improvement. 131-138 2014 SLTU https://www.isca-archive.org/sltu_2014/tarjan14_sltu.html conf/sltu/2014 db/conf/sltu/sltu2014.html#TarjanFM14 Péter Mihajlik András Balog Lightly supervised acoustic model training for imprecisely and asynchronously transcribed speech. 1-5 2013 SpeD https://doi.org/10.1109/SpeD.2013.6682653 conf/sped/2013 db/conf/sped/sped2013.html#MihajlikB13 Balázs Tarján Gellért Sárosi Tibor Fegyó Péter Mihajlik Improved recognition of Hungarian call center conversations. 1-6 2013 SpeD https://doi.org/10.1109/SpeD.2013.6682652 conf/sped/2013 db/conf/sped/sped2013.html#TarjanSFM13 Balázs Tarján Tamás Mozsolics András Balog D. Halmos Tibor Fegyó Péter Mihajlik Broadcast news transcription in Central-East European languages. 59-64 2012 CogInfoCom https://doi.org/10.1109/CogInfoCom.2012.6421940 conf/coginfocom/2012 db/conf/coginfocom/coginfocom2012.html#TarjanMBHFM12 Gellért Sárosi Balázs Tarján András Balog Tamás Mozsolics Péter Mihajlik Tibor Fegyó On modeling non-word events in Large Vocabulary Continuous Speech Recognition. 649-653 2012 CogInfoCom https://doi.org/10.1109/CogInfoCom.2012.6421932 conf/coginfocom/2012 db/conf/coginfocom/coginfocom2012.html#SarosiTBMMF12 Éva Székely Tamás Gábor Csapó Bálint Tóth Péter Mihajlik Julie Carson-Berndsen Synthesizing expressive speech from amateur audiobook recordings. 297-302 2012 SLT https://doi.org/10.1109/SLT.2012.6424239 conf/slt/2012 db/conf/slt/slt2012.html#SzekelyCTMC12 Gellért Sárosi Mihály Mozsáry Péter Mihajlik Tibor Fegyó Comparison of feature extraction methods for speech recognition in noise-free and in traffic noise environment. 1-8 2011 SpeD https://doi.org/10.1109/SPED.2011.5940729 conf/sped/2011 db/conf/sped/sped2011.html#SarosiMMF11

László Tóth 0001 Balázs Tarján Gellért Sárosi Péter Mihajlik Speech Recognition Experiments with Audiobooks. 695-713 2010 19 Acta Cybern. 4 https://cyber.bibl.u-szeged.hu/index.php/actcybern/article/view/3792 db/journals/actaC/actaC19.html#TothTSM10

Péter Mihajlik Zoltán Tüske Balázs Tarján Bottyán Németh Tibor Fegyó Improved Recognition of Spontaneous Hungarian Speech - Morphological and Acoustic Modeling Techniques for a Less Resourced Task. 1588-1600 2010 18 IEEE Trans. Speech Audio Process. 6 https://doi.org/10.1109/TASL.2009.2038807 db/journals/taslp/taslp18.html#MihajlikTTNF10

Gellért Sárosi Tamás Mozsolics Balázs Tarján András Balog Péter Mihajlik Tibor Fegyó Recognition of Multiple Language Voice Navigation Queries in Traffic Situations. 199-213 2010 COST 2102 Conference https://doi.org/10.1007/978-3-642-25775-9_20 conf/cost/2010 db/conf/cost/cost2010.html#SarosiMTBMF10 Balázs Tarján Péter Mihajlik On morph-based LVCSR improvements. 10-16 2010 SLTU https://www.isca-archive.org/sltu_2010/tarjan10_sltu.html conf/sltu/2010 db/conf/sltu/sltu2010.html#TarjanM10 Péter Mihajlik Balázs Tarján Zoltán Tüske Tibor Fegyó Investigation of morph-based speech recognition improvements across speech genres. 2687-2690 2009 INTERSPEECH https://doi.org/10.21437/Interspeech.2009-125 conf/interspeech/2009 db/conf/interspeech/interspeech2009.html#MihajlikTTF09 Géza Németh Gábor Olaszy Mátyás Bartalis Géza Kiss Csaba Zainkó Péter Mihajlik Csaba Haraszti Automated Drug Information System for Aged and Visually Impaired Persons. 238-241 2008 ICCHP https://doi.org/10.1007/978-3-540-70540-6_35 conf/icchp/2008 db/conf/icchp/icchp2008.html#NemethOBKZMH08 Péter Mihajlik Tibor Fegyó Zoltán Tüske Pavel Ircing A morpho-graphemic approach for the recognition of spontaneous speech in agglutinative languages - like Hungarian. 1497-1500 2007 INTERSPEECH https://doi.org/10.21437/Interspeech.2007-433 conf/interspeech/2007 db/conf/interspeech/interspeech2007.html#MihajlikFTI07 Géza Németh Gábor Olaszy Mátyás Bartalis Géza Kiss Csaba Zainkó Péter Mihajlik Speech based drug information system for aged and visually impaired persons. 2533-2536 2007 INTERSPEECH https://doi.org/10.21437/Interspeech.2007-673 conf/interspeech/2007 db/conf/interspeech/interspeech2007.html#NemethOBKZM07 Péter Mihajlik Tibor Fegyó Bottyán Németh Zoltán Tüske Viktor Trón Towards Automatic Transcription of Large Spoken Archives in Agglutinating Languages - Hungarian ASR for the MALACH Project. 342-349 2007 conf/tsd/2007 TSD https://doi.org/10.1007/978-3-540-74628-7_45 db/conf/tsd/tsd2007.html#MihajlikFNTT07 Zoltán Tüske Péter Mihajlik Zoltán Tobler Tibor Fegyó Robust voice activity detection based on the entropy of noise-suppressed spectrum. 245-248 2005 INTERSPEECH https://doi.org/10.21437/Interspeech.2005-119 conf/interspeech/2005 db/conf/interspeech/interspeech2005.html#TuskeMTF05 Péter Mihajlik Zoltán Tobler Zoltán Tüske Géza Gordos Evaluation and optimization of noise robust front-end technologies for the automatic recognition of Hungarian telephone speech. 2677-2680 2005 INTERSPEECH https://doi.org/10.21437/Interspeech.2005-258 conf/interspeech/2005 db/conf/interspeech/interspeech2005.html#MihajlikTTG05 Tibor Fegyó Péter Mihajlik Péter Tatai Comparative study on hungarian acoustic model sets and training methods. 2003 INTERSPEECH https://doi.org/10.21437/Eurospeech.2003-186 conf/interspeech/2003 db/conf/interspeech/interspeech2003.html#FegyoMT03 829-832 Tibor Fegyó Péter Mihajlik Máté Szarvas Péter Tatai Gábor Tatai Voxenter^TM - intelligent voice enabled call center for hungarian. 2003 INTERSPEECH https://doi.org/10.21437/Eurospeech.2003-276 conf/interspeech/2003 db/conf/interspeech/interspeech2003.html#FegyoMSTT03 1905-1908 Tibor Fegyó Péter Mihajlik Péter Tatai Géza Gordos Pronunciation modeling in hungarian number recognition. 1465-1468 2001 INTERSPEECH https://doi.org/10.21437/Eurospeech.2001-30 conf/interspeech/2001 db/conf/interspeech/interspeech2001.html#FegyoMTG01

Máté Szarvas Tibor Fegyó Péter Mihajlik Péter Tatai Automatic Recognition of Hungarian: Theory And Practice. 237-251 2000 3 Int. J. Speech Technol. 3-4 https://doi.org/10.1023/A%3A1026515132762 db/journals/ijst/ijst3.html#SzarvasFMT00

Shaimaa Alwaisi Ákos Antal András Balog Piroska Barta Mátyás Bartalis Csaba Bordás Julie Carson-Berndsen Tamás Gábor Csapó Gergely Dobsinszki A. Seza Dogruöz Gergely Dosinszky Tibor Fegyó Péter Földesy Fruzsina Sára Fruzsina Máté Gedeon Géza Gordos Tekla Etelka Gráczi D. Halmos Csaba Haraszti Amin Hassairi Pavel Ircing Kinga Jelencsik-Mátyus Mate KadarMate S. KadarMáté Soma Kádár Gábor Kiss Géza Kiss Anna Kohári Gernot Kubin Julian Linke Yue Luo Katalin Mády Yan Meng Dalai Mengke Mihály Mozsáry Tamás Mozsolics Bottyán Németh Géza Németh Gábor Olaszy Krisztina Rábai Gellért Sárosi Barbara Schuppler Lili Szabó Máté Szarvas György Szaszák Éva Székely Balázs Tarján Gábor Tatai Péter Tatai Zoltán Tobler Bálint Tóth László Tóth 0001 Viktor Trón Zoltán Tüske Ádam Varga Wuyerigen Xihang Xia Csaba Zainkó