جست‌وجوی استاد و پژوهشگر

• Speech Recognition
• Speaker Verification
• Language Identification
• +۷ مورد دیگر

Professor Jan Černocký serves as Head of Department at the Department of Computer Graphics and Multimedia (DCGM) within the Faculty of Information Technology at Brno University of Technology (FIT VUT). With a professional email cernocky@fit.vut.cz and office L221.2, he maintains an active research profile with numerous publications spanning over 20 years in the field of speech processing and recognition. His work is well-documented through multiple research identifiers including ORCID iD 0000-0002-8800-0210, Scopus Author ID 6604040821, and Researcher ID M-7494-2019. Professor Černocký's research interests focus primarily on advanced speech processing technologies, with particular emphasis on speech recognition, speaker verification, language identification, and multimodal systems. His work demonstrates a strong trajectory from traditional speech processing techniques toward modern deep learning approaches, especially in self-supervised learning for speech applications. Recent publications show his leadership in developing benchmarks like TS-SUPERB for target speech processing and innovative methods for speaker verification using transformer models. His research group at BUT has made significant contributions to multi-channel speech processing, target speech extraction, and speaker diarization systems. The analysis of Professor Černocký's recent publications (2023-2025) reveals several key trends in his research direction. There's a clear shift toward self-supervised learning approaches for speech processing, with numerous papers exploring how pre-trained models can be adapted for speaker verification, target speech extraction, and multi-channel processing. His work increasingly incorporates transformer architectures and attention mechanisms, reflecting the broader trends in speech processing research. The 2024 publications particularly highlight work on multimodal analysis (BESST dataset for stress detection) and practical applications of speech technology for social inclusion. Throughout his career, Professor Černocký has maintained strong collaborative relationships with researchers across Europe and internationally, evidenced by his extensive publication record with co-authors from multiple institutions. His leadership role as Head of Department at DCGM places him at the center of speech processing research at Brno University of Technology, where his team continues to produce cutting-edge research in speech technology.

• Natural Language Processing
• Artificial Intelligence
• Information Extraction
• +۴ مورد دیگر

Assoc. Prof. Pavel Pecina, Ph.D., is a faculty member at the Institute of Formal and Applied Linguistics, Faculty of Mathematics and Physics, Charles University, Prague, Czech Republic. His primary academic role as an Associate Professor focuses on Natural Language Processing , Artificial Intelligence , and related areas. Research Interests: Information extraction, information retrieval, machine translation, multimodal data interpretation, optical music recognition Teaching: Courses since 2012-2022 including Natural Language Processing , Information Retrieval , and Statistical Methods in NLP Key Research Projects include GI-Insight (AI for healthcare), RES-Q+ (global stroke care registry), and MEMORISE (heritage digitization). His recent publications (2023-2025) demonstrate expertise across clinical NLP , historical document analysis , and music technology domains. Advising: Supervised 10+ graduate students in areas ranging from semantic search to optical music recognition. Collaborates with international institutions on multilingual translation and medical informatics challenges through initiatives like IWSLT and CLEF eHealth.

• interpretation of deep neural networks
• machine learning
• dependency syntax
• +۳ مورد دیگر

David Mareček is a researcher at the Institute of Formal and Applied Linguistics (ÚFAL) within the Faculty of Mathematics and Physics at Charles University. His work focuses on neural network interpretation, machine learning, dependency syntax analysis, and machine translation, with a particular emphasis on unsupervised and semi-supervised parsing techniques. Current projects: GenderBias (2023–present), EduPo (2024–present) Long-term contributions in dependency treebanks (HamleDT, 2012–2016) and neural machine translation frameworks (Treex, TectoMT) His research explores linguistic structure in neural architectures, including BERT and Transformer models, through methods like structural probing and attention analysis. Recent publications examine gender bias mitigation, automated poetry analysis, and theatre script generation. Key collaborations include Rudolf Rosa, Tomáš Musil, and Tomasz Limisiewicz. He supervises students in projects spanning multilingual representations, language model probing, and bias detection in AI systems. Technical leadership in developing tools like THEaiTRobot and DEPFIX demonstrates his commitment to applied computational linguistics and robust NLP systems.

Milan Strakaمشاهده پروفایل

استاد

• Machine Learning
• Artificial Neural Networks
• Deep Learning
• +۱۲ مورد دیگر

Milan Straka is a professor at the Institute of Formal and Applied Linguistics (ÚFAL) within the Faculty of Mathematics and Physics at Charles University, Prague. His research focuses on machine learning, artificial neural networks, deep learning, and structured prediction with applications to natural language processing (NLP) tasks including POS tagging, dependency parsing, named entity recognition, and optical music recognition. University: Charles University School: Faculty of Mathematics and Physics Department: Institute of Formal and Applied Linguistics Straka has developed several NLP tools like UDPipe and NameTag, contributing to multilingual coreference resolution, Czech grammar error correction, and historical language processing pipelines. His recent work involves contextualized embeddings, semantic parsing, and large-scale dataset creation for Czech NLP tasks. He collaborates on European Language Grid initiatives and has participated in multiple shared tasks (CRAC, MRP, W-NUT). His publications span topics from algorithm design (functional data structures) to modern transformer-based language models (RoBERTa) with emphasis on Czech language resources.

• Neural Machine Translation
• Multilingual Neural Machine Translation
• Simultaneous Speech Translation
• +۴ مورد دیگر

Dominik Macháček is a researcher at the Institute of Formal and Applied Linguistics (ÚFAL) at Charles University's Faculty of Mathematics and Physics in Prague. He successfully defended his PhD thesis titled 'Multi-Source Simultaneous Speech Translation' in June 2024 and is currently active in research while preparing for a post-doctoral fellowship at the University of Edinburgh starting in January 2026. His research focuses on Neural Machine Translation , particularly Simultaneous Speech Translation where translation happens in real-time with minimal latency (typically 2-4 seconds). His work spans Multilingual Neural Machine Translation, Quality Estimation for speech translation systems, and Human-to-Human Cross-Lingual Communication. His current 'Live Credible Translation' project aims to enhance simultaneous speech translation by adding quality estimation modules to improve credibility of real-time translations. Macháček has secured a US patent (12,056,457 B2) in August 2024 for a 'Computer-implemented method of real time speech translation and a computer system for carrying out the method'. His research has been presented at major conferences including ACL and IWSLT, particularly through his work on the SimulStreaming system which enables large foundation models like Whisper to operate in simultaneous mode. His publication record shows a consistent focus on real-time translation challenges, with recent work examining robustness to transcription errors, correlation between automatic metrics and human ratings, and continuous rating methodologies for human evaluation. His research trajectory demonstrates progression from theoretical foundations to practical applications with significant industry relevance. Macháček actively supervises Bachelor's and Master's theses and teaches the NPFL140 Large Language Models course. He is involved in multiple research projects including CEDMO 2.0, InCroMin, Sign LLM, NEUREM3, and ELITR. His future work will focus on advancing the state of the art in simultaneous speech translation through his 'Live Credible Translation' project at the University of Edinburgh.