Projects Activity

• ENFIELD: European Lighthouse to Manifest Trustworthy and Green AI.
  • 2023-2026, Researcher, European Horizon 2020 Commission (doi)
• EKÖP: Egyetemi Kutatói Ösztöndíj Program.
  • Sep 2024 - Aug 2025, Principal Investigator, National Research, Development and Innovation Office
• AI4Europe: Building a European AI on-demand platform.
  • 2019-2022, Researcher, European Horizon 2020 Commission (doi)
• APH-ALARM: Comprehensive safety solution for people with Aphasia.
  • 2020-2023, Researcher, AAL-Europe (url)
• OTKA-FK-124584: Silent Speech Interface based on articulatory movements.
  • 2017-2022, Investigator, Hungarian Fund (url)

Supervising

• Main Supervisor
  • PhD students
  • Ali Mandeel, Iraqi (2024 - 2025)
  • Rammi Kammoun, Tunisian (2024 – 2028)
  • MSc student
  • Zineb Hammadi, Algerian (2023 – 2025)
  • Riad Larbi, Algerian (2023 – 2025)
  • Layan Sawalha, Jordanian (2021 – 2023, thesis)
  • BSc students
  • Felipe Lopes Franklin Bezerra, Brazilian (2021 – 2022, thesis)
  • Suciu Barnabás, Hungarian (2020 – 2021, thesis)
  • BSc & MSc Project laboratories (2020 – present)
• Co-Supervisor
  • PhD students
  • Ali Mandeel, Iraqi (2020 - 2023)
  • Shaimaa S. Alwaisi, Iraqi (2022 - 2026)
  • Ali Abduljabbar Jawad, Iraqi (2024 - 2028), PhD Research Plan, Polytechnic University of Catalonia (UPC), Barcelona, Spain
  • MSc students
  • Dean Reymen, Belgian (2021 – 2022), thesis in Hasselt University, Belgium
  • Pengyu Dai, Chinese (2019 – 2021, thesis)
• Main Examiner for Project Laboratories
  • Responsible for assessing and assigning grades to BSc and MSc student projects

Teaching

• Deep learning (theoretical lecture) + (practical lecture)
  • Introduction and Software Stack [slides]
  • Backpropagation [slides]
  • Hardware Architecture [slides]
  • Hyperparameter Optimization [slides]
  • Frameworks, Pytorch basics [slides]
  • Classification and CNN [slides]
  • Transfer Learning and GANs [slides]
  • Computer Vision [slides]
  • Automatic Speech Recognition [slides]
  • Speech Technology and TTS [slides]
  • Graph Neural Networks [slides]
• Info-communications (theoretical lecture)
  • Sound and Hearing [slides]
  • Image and Vision [slides]
  • Speech Processing [slides]
  • Radio communication [slides]
  • Video broadcasting [slides]
• Human-Computer Interaction (practical lecture)
  • User interface
  • Dialogue systems
  • Case studies
• Smart City Laboratory (lab)
  • Augmented reality
  • Virtual reality
  • AR frameworks

Full Publications

• 2025
1. Mohammed Salah Al-Radhi, Géza Németh, Breaking the Silence: Brain-to-Speech Innovations, Large-Scale Data Analytics, European Research Consortium for Informatics and Mathematics (ERCIM), No. 140, 2025.

   [paper]    

2. Botond Bárdos-Deák, Adrián Tibor Bodai, Mohammed Salah Al-Radhi, Optimizing Cardiac MRI Segmentation: An Ensemble Approach with U-Net Variants, 3rd Workshop on Intelligent Infocommunication Networks, Systems and Services (WINS), Budapest, Hungary, 2025.

   [paper]    [poster]

3. Riad Larbi, Mohammed Salah Al-Radhi, Architectural Enhancements and Feature Optimization of AutoVocoder for High-Quality Speech Synthesis, 3rd Workshop on Intelligent Infocommunication Networks, Systems and Services (WINS), Budapest, Hungary, 2025.

   [paper]    [slides]

4. Zineb Hammadi, Mohammed Salah Al-Radhi, A Hybrid Algorithm for Robust Pitch Estimation in Emotional Speech Synthesis, 3rd Workshop on Intelligent Infocommunication Networks, Systems and Services (WINS), Budapest, Hungary, 2025.

   [paper]    [poster]

• 2024
1. Rami Kammoun, Mohammed Salah Al-Radhi, Géza Németh, Enhancing Expressive TTS Synthesis for Multilingual Low-Resource Languages: Challenges and Applications, Acoustics and Speech Processing Conference, Bratislava, Slovakia, 2024.

   [paper]    [slides]

2. Shaimaa Alwaisi, Mohammed Salah Al-Radhi, Géza Németh, ChildTinyTalks (CTT): A Benchmark Dataset and Baseline for Expressive Child Speech Synthesis, 26th International Conference on Speech and Computer (SPECOM). Lecture Notes in Computer Science, Belgrade, Serbia, 2024.

   [paper]    [slides]

3. Shaimaa Alwaisi, Mohammed Salah Al-Radhi, Géza Németh, Multi-Speaker Child Speech Synthesis in Low-Resource Hungarian Language, 2nd Workshop on Intelligent Infocommunication Networks, Systems and Services (WINS), Budapest, Hungary, 2024.

   [paper]    [slides]

4. Shukhrat Kulboboev, Mohammed Salah Al-Radhi, Improving Speech Naturalness and Nuance using HiFiGAN-Hubert-Soft Vocoder: A Case Study of the Voicebox TTS Model, 2nd Workshop on Intelligent Infocommunication Networks, Systems and Services (WINS), Budapest, Hungary, 2024.

   [paper]    [slides]

• 2023
1. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Non-Parallel Voice Conversion with Emphasis on Expressive Voice: A Style-Based Approach, Socially responsible–Applied Linguistics Conference, p. 19, Budapest, Hungary, 2023.

   [proceedings]    [certificate]

2. Peter Mayer, Katharina Werner, Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Bálint Czeba, Géza Németh, Ana Patrícia Rocha, Ilídio C. Oliveira, Samuel Silva, Melinda Szeker, António Teixeira, Paul Panek, Concept and Pictogram-Based User-Interface Design of a Helper Tool for People with Aphasia, 17th Annual Conference on Health Informatics meets Digital Health (dHealth), vol 301, pp. 77-82, Vienna, Austria, 2023.

   [paper]    [project]

3. Shaimaa Alwaisi, Mohammed Salah Al-Radhi, Géza Németh, Universal Approach to Multilingual Multispeaker Child Speech Synthesis, 12th ISCA Speech Synthesis Workshop (SSW), pp. 236-237, Grenoble, France, 2023.

   [paper]    [poster]

4. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Nonparallel Expressive TTS for Unseen Target Speaker using Style-Controlled Adaptive Layer and Optimized Pitch Embedding, 12th International Conference on Speech Technology and Human-Computer Dialogue (SpeD), pp. 176-181, Bucharest, Romania, 2023.

   [paper]    [slides]    [demo]

5. Ali Raheem Mandeel, Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Modeling Irregular Voice in End-to-End Speech Synthesis via Speaker Adaptation, 12th International Conference on Speech Technology and Human-Computer Dialogue (SpeD), pp. 170-175, Bucharest, Romania, 2023.

   [paper]    [demo]

6. Shaimaa Alwaisi, Mohammed Salah Al-Radhi, Géza Németh, Automated Child Voice Generation: Methodology and Implementation, 12th International Conference on Speech Technology and Human-Computer Dialogue (SpeD), pp. 48-53, Bucharest, Romania, 2023.

   [paper]    [demo]

7. Ali Raheem Mandeel, Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Enhancing End-to-End Speech Synthesis by Modeling Interrogative Sentences with Speaker Adaptation, 12th International Conference on Speech Technology and Human-Computer Dialogue (SpeD), pp. 158-163, Bucharest, Romania, 2023.

   [paper]    [demo]

8. Mohammed Salah Al-Radhi, Omnia Ibrahim, Ali Raheem Mandeel, Tamás Gábor Csapó, Géza Németh, Advancing Limited Data Text-to-Speech Synthesis: Non-Autoregressive Transformer for High-Quality Parallel Synthesis, 12th International Conference on Speech Technology and Human-Computer Dialogue (SpeD), pp. 152-157, Bucharest, Romania, 2023.

   [paper]    [slides]    [demo]

9. Layan Swalha, Mohammed Salah Al-Radhi, Improving Naturalness of Neural-based TTS System Trained with Limited Data, 1st Workshop on Intelligent Infocommunication Networks, Systems and Services (WINS), pp. 71-75, Budapest, Hungary, 2023.

   [paper]    [poster]

10. Ali Raheem Mandeel, Ammar Abdullah Aggar, Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Implementing a Text-to-Speech synthesis model on a Raspberry Pi for Industrial Applications, 1st Workshop on Intelligent Infocommunication Networks, Systems and Services (WINS), pp. 77-81, Budapest, Hungary, 2023.

    [paper]    [poster]

11. Ali Raheem Mandeel, Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Investigations on speaker adaptation using a continuous vocoder within recurrent neural network based textto-speech synthesis, Multimedia Tools and Applications, 82, pp. 15635–15649, 2023.

    [paper]

12. Ali Raheem Mandeel, Ammar Abdullah Aggar, Mohammed Salah Al-Radhi, Tamás Gábor Csapó, A Smart Control System for the Oil Industry Using Text-to-Speech Synthesis Based on IIoT, Electronics, 12(16):3380, 2023.

    [paper]

13. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Improving the expressiveness of TTS synthesis with non-autoregressive neural vocoding, Beszédkutatás–Speech Research, pp. 94-96, Budapest, Hungary, 2023.

    [proceedings]    [poster]

14. Ali Raheem Mandeel, Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Creaky Voice via Speaker Adaptation within End-to-End Text to Speech Synthesis, Beszédkutatás–Speech Research, pp. 78-80, Budapest, Hungary, 2023.

    [proceedings]    [slides]

15. Layan Sawalha, Mohammed Salah Al-Radhi, Few-Shot Multi-Language Text-to-Speech Synthesis with State-of-the-Art Neural Networks, Beszédkutatás–Speech Research, pp. 97-99, Budapest, Hungary, 2023.

    [proceedings]

• 2022
1. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Csaba Zainkó, Géza Németh, Towards Parametric Speech Synthesis Using Gaussian-Markov Model of Spectral Envelope and Wavelet-Based Decomposition of F0, 30th European Signal Processing International Conference (EUSIPCO), Belgrade, Serbia, pp. 1150–1154, 2022.

   [paper]    [slides]    [demo]

2. Ali Raheem Mandeel, Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Speaker Adaptation Experiments with Limited Data for End-to-End Text-To-Speech Synthesis using Tacotron2, Infocommunications Journal, XIV, 3, pp. 55–62, 2022.

   [paper]    [demo]

3. Ismaeil R. Alnaab, Harwan M. Taha, Zainab A. Abdulwahab, Mohammed Salah Al-Radhi, Performance comparison between fixed tilt angle and solar tracking systems at Basra governorate: A case study, Indonesian Journal of Electrical Engineering and Computer Science, 26(1):184–193, 2022.

   [paper]

4. Safa Jameel Al-Kamil, Mohammed Salah Al-Radhi, Deep Learning for Self-Driving Vehicles, 2nd International Multi-Disciplinary Conference Theme: Integrated Sciences and Technologies (IMDC-IST), Sakarya, Turkey, 2022.

   [paper]

• 2021
1. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Csaba Zainkó, Géza Németh, Continuous Wavelet Vocoder-Based Decomposition of Parametric Speech Waveform Synthesis, in Proceedings of Interspeech Conference, Brno, Czechia, pp. 2212-2216, 2021.

   [paper]    [slides]    [demo]

2. Pengyu Dai, Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Effects of F0 Estimation Algorithms on Ultrasound-Based Silent Speech Interfaces, 11th International Conference on Speech Technology and Human-Computer Dialogue (SpeD), Bucharest, Romania, pp. 47-51, 2021.

   [paper]    [slides]

3. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Noise and acoustic modeling with waveform generator in text-to-speech and neutral speech conversion, Multimedia Tools and Applications, 80, 1969–1994, 2021.

   [paper]

4. Ali Raheem Mandeel, Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Speaker Adaptation with Continuous Vocoder-Based DNN-TTS, 23th International Conference on Speech and Computer (SPECOM). Lecture Notes in Computer Science, 12997, pp. 407–416, 2021.

   [paper]

5. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Effects of Sinusoidal Model on Non-Parallel Voice Conversion with Adversarial Learning, Applied Sciences, 11, 7489, pp. 1–16, 2021.

   [paper]    [demo]

6. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Advances in speech vocoding for text-to-speech with continuous parameters, 2nd International Conference on Artificial Intelligence and Speech Technology (AIST), Delhi, India, 2021.

   [chapter]

• 2020
1. Mohammed Salah Al-Radhi, High-Quality Vocoding Design with Signal Processing for Speech Synthesis and Voice Conversion, PhD Dissertation, Budapest University of Technology and Economics, Faculty of Electrical Engineering and Informatyics, 2020.

   [thesis]

2. Mohammed Salah Al-Radhi, Omnia Abdo, Tamás Gábor Csapó, Sherif Abdou, Géza Németh, Mervat Fashal, A continuous vocoder for statistical parametric speech synthesis and its evaluation using an audio-visual phonetically annotated Arabic corpus, Computer Speech and Language, Volume 60, pp. 1-15, 2020.

   [paper]

3. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Continuous Noise Masking Based Vocoder for Statistical Parametric Speech Synthesis, IEICE Transactions on Information and Systems, E103.D, Issue 5, pp. 1099–1107, 2020.

   [paper]    [demo]

4. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, conTTS: Text-toSpeech Application using a Continuous Vocoder, 12th International Seminar on Speech Production (ISSP), pp. 170-173, New Haven CT, USA, 2020.

   [proceedings]    [demo]

5. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Non-Parallel Voice Conversion Incorporating Sinusoidal Model with Adversarial Learning, Beszédkutatás–Speech Research, pp. 77-79, 2020.

   [paper]

• 2019
1. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Continuous vocoder applied in deep neural network based voice conversion, Multimedia Tools and Applications, 78, pp. 33549–33572, 2019.

   [paper]    [demo]

2. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Parallel Voice Conversion Based on a Continuous Sinusoidal Model, 10th International Conference on Speech Technology and Human-Computer Dialogue (SpeD), Timisoara, Romania, pp. 1-6, 2019.

   [paper]    [slides]    [demo]

3. Tamás Gábor Csapó, Mohammed Salah Al-Radhi, Géza Németh, Gábor Gosztolya, Tamás Grósz, László Tóth, Alexandra Markó, Ultrasound-Based Silent Speech Interface Built on a Continuous Vocoder, in Proceedings of Interspeech Conference, Graz, Austria, pp. 894-898, 2019.

   [paper]    [demo]

4. Waleed I., Baha Sawadi, Safa Al-Kamil,Mohammed Salah Al-Radhi, Yasir Al-Yasir, Ameer Saleh, Raed Abd-Alhameed, Prediction of Solar Irradiance Based on Artificial Neural Networks, Inventions, 4(3), 45, pp. 1-10, 2019.

   [paper]

5. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Adaptive Refinements of Pitch Tracking and HNR Estimation within a Vocoder for Statistical Parametric Speech Synthesis, Applied Sciences, 9 (12), 2460, pp. 1–23, 2019.

   [paper]    [demo]

6. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, RNN-based speech synthesis using a continuous sinusoidal model, International Joint Conference on Neural Networks (IJCNN), Budapest, Hungary, pp. 1-8, 2019.

   [paper]    [slides]    [demo]

7. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, High quality continuous vocoder in deep recurrent neural network based speech synthesis, in Eastern European Machine Learning, Google DeepMind, Bucharest, Romania, 2019.

   [poster]

• 2018
1. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, A Continuous Vocoder using Sinusoidal Model for Statistical Parametric Speech Synthesis, 20th International Conference on Speech and Computer (SPECOM), Lecture Notes in Computer Science, Leipzig, Germany, pp. 11–20, 2018.

   [paper]    [slides]    [demo]

2. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Improving continuous F0 estimator with adaptive time-warping for high-quality speech synthesis, in Beszédkutatás (conference of the speech reseacrch), Budapest, Hungary, pp. 77-79, 2018.

   [proceedings]

• 2017
1. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Time-Domain Envelope Modulating the Noise Component of Excitation in a Continuous Residual-Based Vocoder for Statistical Parametric Speech Synthesis, In proceedings of Interspeech, Stockholm, Sweden, pp. 434-438, 2017.

   [paper]    [slides]    [demo]

2. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Continuous vocoder in feed-forward deep neural network based speech synthesis, International conference of digital speech and image processing, Novi Sad, Serbia, pp. 1-4, 2017.

   [paper]    [slides]    [demo]

3. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Effects of adding a Harmonic-to-Noise Ratio parameter to a continuous vocoder, in Proceedings of the 6th of the UK Speech, Cambridge University, England, 2017.

   [poster]

4. Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh, Deep Recurrent Neural Networks in Speech Synthesis Using a Continuous Vocoder, 19th International Conference on Speech and Computer (SPECOM). Lecture Notes in Computer Science, Hatfield, UK, pp. 282–291, 2017.

   [paper]    [poster]

5. Mohammed Salah Al-Radhi, High quality continuous residual-based vocoder for statistical parametric speech synthesis, 3rd Doctoral Consortium, International Speech Communication Association (ISCA-SAC), Interspeech, KTH Royal Institute of Technology, Stockholm, Sweden, 2017.

   [paper]

• 2012
1. Mohammed Salah Al-Radhi, Design of Finite Impulse Response Digital Filters using Optimal Methods, MSc Dissertation, Portsmouth University, School of Energy and Electronic Engineering, Portsmouth, England, 2012.

   [thesis]

2. Mohammed Salah Al-Radhi, Performance of Convolutional coding with Hard decision Viterbi decoding on BPSK Systems over Noisy Channels, Student Symposium, Portsmouth University, School of Engineering, Portsmouth, England, 2012.

   [paper]