Postdoctoral Researcher in Speech and Language Technology

At Aalto University, science and art meet technology and economics. We build a sustainable future by achieving breakthroughs in our key areas and their intersections. At the same time, we inspire the changemakers of the future and create solutions to the world's great challenges. Our university community includes 16,000 students, 446 professors, and over 5,000 employees, representing a total of over 120 nationalities. Our campus is located in Espoo, Otaniemi. Diversity is part of us, and we work continuously to ensure the diversity and non-discrimination of our community. Therefore, we encourage qualified applicants from all backgrounds to join our community.

Research is conducted at the Aalto University Department of Information and Telecommunication Engineering, in DICE. The project environment offers excellent facilities for research in deep learning, speech, and audio. Available resources include, among others, Aalto University's scientific computing cluster, CSC's national computing infrastructure including LUMI, and the Aalto Acoustics Laboratory's anechoic chambers, listening rooms, and audio measurement equipment.

We are now looking for a postdoctoral researcher for the field of speech and language technology.

Are you interested in transparency, content originality, and traceability in AI-generated speech, music, and audio? We are looking for a postdoctoral researcher to join the COWAMA: Content-based watermarking for AI-generated audio project, led by Assistant Professor Lauri Juvela.

Generative AI is currently capable of producing increasingly realistic speech and music, which raises important questions regarding misinformation, copyrights, ownership, attribution, and responsibility. The EU AI Act requires that generative AI systems mark the produced content, but external metadata is easy to remove. Therefore, technical methods are also needed to support transparency and identification. In this project, watermarking methods and software will be developed for generative models of speech and audio. Particular focuses are the robustness of the methods and their suitability for open-source use.

Your Role

As a postdoctoral researcher, you will take primary responsibility for developing deep content-based watermarks for audio and will work together with a doctoral researcher on robustness, evaluation, and generalizability.

Your tasks include:

• Developing new content-based watermarking methods for generated speech, music, and audio
• Designing detection methods and quality metrics for situations where generated audio may deviate from the original reference content
• Developing methods for, for example, audio content identification, music information retrieval, audio fingerprinting, feature correspondences in self-supervised audio representations, and reference-free audio quality prediction
• Integrating watermarking into deep generative audio models, including the use of collaborative watermarking in diffusion models and audio language models based on neural audio codecs
• Investigating whether watermarking information can be embedded into long-term content, such as speech semantics, duration, prosody, or emphasis, and not just local signal-level features
• Developing robustness evaluation methods, differentiable augmentation tools, and realistic attack scenarios for audio watermarks
• Publishing research in leading speech, audio, and machine learning publications and conferences, as well as participating in open releases of software, trained models, and reproducible research outputs.

Research methods include experimental design, software implementation, running deep learning experiments on high-performance computing infrastructure, and evaluation using objective metrics and subjective listening tests.

Your Network and Team

Your supervisor will be Assistant Professor Lauri Juvela, who leads Aalto University's Speech Synthesis research group. The group works on deep generative speech and audio models, speech synthesis, differentiable signal processing, deepfake detection, and watermarking.

In addition to Aalto's speech groups and the project's postdoctoral researcher, you will work in an international collaborative network that includes, among others, Professor Junichi Yamagishi from the National Institute of Informatics in Japan, Professor Xavier Serra from Universitat Pompeu Fabra in Spain, and Professor Gustav Eje Henter from KTH Royal Institute of Technology in Sweden. The collaborative network has a strong background in speech and audio synthesis, deep generative models, and deepfake detection, making the project well-positioned to influence watermarking and source tracing of AI-generated audio.

What We Expect From You

We are looking for a motivated researcher with a strong interest in generative AI, speech, music, and audio technologies. Success in the position requires:

• A PhD or a soon-to-be completed PhD in speech or audio processing, machine learning, signal processing, computer science, or another suitable field
• Strong experience in deep learning and generative models
• Programming skills suitable for modern machine learning research, including Python and deep learning tools
• Interest in one or more of the following areas: speech synthesis, music generation, neural audio codecs, diffusion models, audio language models, watermarking, deepfake detection, audio quality assessment, or learning audio representations
• The ability to conduct independent research and publish in high-level peer-reviewed publication forums
• Motivation to work in an open science-oriented project developing open-source software and reproducible research outputs
• Good collaboration and communication skills in an international research environment
• Fluent English language skills. Proficiency in Finnish is not required.

Experience in doctoral and master's level supervision is considered an advantage.

What We Offer

• A meaningful and timely research topic at the intersection of generative AI, audio technology, transparency, and digital trust. The project addresses the need to develop best practices and standards for handling AI-generated content with open, transparent, and decentralized academic solutions.
• The opportunity to work on methods that improve the transparency and security of AI-generated speech and audio through watermarking, source tracing, and deepfake detection
• Access to excellent computing and research infrastructure, including Aalto's CPU/GPU clusters, CSC and LUMI, FIN-CLARIN's language resources, and the Aalto Acoustics Laboratory
• An international collaborative network of leading researchers and organizations in speech synthesis, music technology, deepfake detection, and audio generation
• Strong support for open science, including open publishing as well as the open release of software, models, and datasets
• A meaningful and inspiring environment. We are proud of our purpose to shape a sustainable future. We renew society with research-based knowledge, creativity, and entrepreneurship.
• A culture where everyone is valued. All our work is guided by the university's values: responsibility, courage, and collaboration. We are an open community where equality and inclusivity enable curiosity, innovation, collaboration, and well-being.
• Support, guidance, and mentoring when needed.
• Great opportunities for skill development and learning. We strive to utilize the strengths and passions of every Aalto person in the best possible way. We help each other succeed, focusing on the community and people learning new things.
• We aim to combine the best aspects of remote and on-site work. Your workplace will be located in the renewed and vibrant Otaniemi campus area, which offers diverse services and good transport links.

The salary for the position is determined according to the university's salary system used at Aalto University. The starting salary for the postdoctoral researcher position is €4,220/month, and the salary increases based on performance. The position is fixed-term and lasts for two years. The project funding has been granted for four years, and it is possible to extend the employment contract for another two years. The position begins in January 2027 or as agreed.

Work tasks must be performed in Finland.

Interested?

If you are excited about the position and want to join us, please send your CV, motivation letter, and copies of degree certificates and transcripts through our recruitment system (at the bottom of the ”Apply now!” page) no later than 31.10.2026 at 23:59 (EET).

Current employees of Aalto University should apply for the position using their own employee profile through the Workday system (internal jobs). Aalto students or visitors apply using a personal email address (not aalto.fi) through the external Open Jobs page.

For more information about the position, contact Lauri Juvela, lauri.juvela@aalto.fi, +358 50 464 6653. For questions regarding the application, HR Advisor Johanna Haapalainen, hr-elec@aalto.fi, will help.

We review applications and may invite suitable applicants for an interview during the application period. You will hear from us by 15.11.2026 at the latest. We strive for a transparent and equal recruitment process, so you can ask us for feedback.

Do you want to know more about us and your future colleagues? You can watch these videos:

This is Aalto University! Aalto University – Towards a better world and Shaping a Sustainable Future.

Read more about working at Aalto: https://www.aalto.fi/fi/toihin-aaltoon and visit the new virtual campus experience: https://virtualtour.aalto.fi.