IEEE International Workshop on
Machine Learning for Signal Processing (MLSP) 2025
August 31-September 3, Istanbul/Turkey
Signal Processing in the age of
Large Language Models
IEEE

INDUSTRY PROGRAM

Industry Panel

September 3rd, 13.30-15.00 (Marmara Room)
We will have the chance to have the perspective of industrial researchers. We will have five prominent industrial researchers to discuss about their experiences in industry and to have them contrast with academic research. This will be an interactive session, so do not forget to bring your questions

Moderator:

Ceyhun B. Akgul

Panelists: 

Prof. Aytul Ercil: Professor Dr. Aytül Erçil received her BSc in Electrical Engineering and Mathematics from Boğaziçi University in 1979, her MSc in Applied Mathematics from Brown University in 1980, and her PhD in 1983. After working at General Motors Research Laboratory for five years, she worked as a faculty member at Boğaziçi University between 1988 and 2001 and as the director of the BUPAM Machine Vision Laboratory, which she founded. Between 2001 and 2013, she worked as a faculty member and the director of the VPALAB Machine Vision Laboratory at Sabancı University, which she founded. Prof. Erçil, who has served as a researcher/project manager in various international projects (NATO, FP4, Eureka, NSF, FP6, Nedo, FP7), is the founding president of TÖTİAD - Turkish Association for Pattern Recognition and Image Analysis, and has served as a board member of the IAPR ‘International Association of Pattern Recognition'. Vistek ISRA Vision Inc., founded by Prof. Erçil in 2006, was sold to the German company ISRA Vision, which is number one in Europe and third in the world in machine vision, in December 2013. Erçil's work has been deemed worthy of many national and international awards such as the 'International Success Award', the Eureka Success Story, the Endeavor entrepreneur, the 2010 technology award finalist, the Veuve Clicquot 'most influential woman entrepreneur of the year' award, the first prize in the Machinery and Accessories Production Technologies competition, the '2013 Turkey's woman entrepreneur', the 'Kristal Ağaç woman entrepreneur of the year', the 2014 Ansiad 'woman entrepreneur of the year', the 'Microsoft Woman Leader Making a Difference in Informatics' award, the 'I-COM Data Startup Challenge' award, the 'Hello Tomorrow Top 500 Startups in Deeptech' award, the 2019 Selçuk Yasar entrepreneurship, innovation, innovation award, the 'EY Startup of the Year Award' IWEC Foundation 2020 Business award, the META Women in Tech award, the IWF 'Women who make a difference' award. She has also been included in Forbes' '50 most powerful women' list for four consecutive years. 

Prof. Erçil serves on the advisory board of the Swiss Innovation Valley and the Scientific and Industrial Advisory Board of the European Machine Vision Association. A serial entrepreneur and certified angel investor, Prof. Erçil is a partner and co-CEO of Vispera Information Technologies Inc.

Prof. Berk Gokberk: Berk Gökberk graduated from the Department of Computer Engineering, Boğaziçi University, Turkey, in 1999. He obtained his MS and PhD degrees from the same department in 2001 and 2006. He focused on 2D and 3D face recognition systems during his graduate study. From 2006 until 2008, he was a senior scientist at the Information Systems and Security department, Philips Research Eindhoven, where he concentrated on computer vision systems and biometrics, especially fingerprint biometrics, biometric template protection, and 3D face recognition systems. From 2008 until 2013, he was with the Signals of Systems group in the Department of Electrical Engineering, Mathematics and Computer Science, University of Twente. From 2014 until 2021, he worked as an Assistant Professor in the Computer Engineering Department at MEF University. Currently, he is a faculty member in the Department of Computer Engineering at Boğaziçi University. Berk Gökberk's research areas include 2D and 3D computer vision systems, biometric systems, the visual analysis of humans, deep learning, machine learning, data analysis, and data visualization. 

Prof. Levent Arslan: Levent Arslan graduated from Bogazici University Electrical and Electronics Engineering department in 1991. Then he received MS and PhD degrees from Duke University Electrical Engineering department in 1993 and 1996. After graduation he worked at Entropic Research, Washington DC as an R&D engineer until 1998.  Between 1998 and 2023 he worked at Bogazici University as a full-time professor. He also started a speech R&D company, Sestek, in 2000. Sestek now has 170 employees, mostly R&D engineers. Sestek offers a wide range of products including conversational AI and Analytics. Prof. Arslan has 21 journal papers, 76 conference papers and 22 patents. His research interests are speech recognition, speech synthesis, voice verification and natural language understanding.

Dr. Ersen Kavak: Dr. Ersen Kavak is an entrepreneur of genomics and a software enthusiast. He is the founder of Genomize Inc. and an adjunct faculty member at Bogazici University. He completed his Ph.D. in Molecular Biology and Genetics at Bogazici University in 2007 and worked at the Ludwig. Institute for Cancer Research, Karolinska Institute, as a Postdoctoral Researcher after that. He published multiple papers on genomic regulation in cancer cells in prestigious journals such as Nature and Nucleic Acid Research. After his time in Karolinska, he founded Genomize, Inc. in 2013 to create a solution that would allow physicians to diagnose patients with rare diseases quickly and reliably. So far, his vision helped physicians to diagnose more than 240.000 patients in more than 25 countries. He also has been teaching Python at Bogazici University for molecular biology students since 2014 and coaching next-generation bioinformaticians.


 

Industrial-Academic Joint Workshop on Emerging Problems and Methods in Audio, Speech and Language Processing

September 1st, 16.00-18.00 (Levent Room 3)
This workshop will cover the perspectives of machine learning researchers in audio and speech. The workshop will cover recent topics such audio representation learning, source separation, explainability, and also industrial perspectives from Huawei. 
 

Agenda

Talks (16:00-17:10)

Each talk: 11 minutes presentation + 3 minutes Q&A

1. Mengyao Zhu (Huawei, China): Challenges and Requirements in the field of Audio from Huawei

2. Zheng-Hua Tan (Aalborg University, Denmark): Emerging Sequence Models for Audio Representation Learning and Speech Enhancement

3. Jinhua Liang (Queen Mary University of London, UK): 

4. Wenwu Wang (University of Surrey, UK): Text-Queried Audio Source Separation

5. Cem Subakan (Laval University/Mila-Quebec AI Institute, Canada): Producing Listenable Explanations for Audio Models


Panel Discussion (17:10-18:00)

Panel Members:

Zheng-Hua Tan, Aalborg University, Denmark

Paris Smaragdis, MIT, USA

Cem Subakan, Laval University/Mila-Quebec AI Institute, Canada

Mengyao Zhu, Huawei, China

Wenwu Wang, University of Surrey, UK

 
For more details regarding the talks please refer to the pdf document here