Postdoc in AI Audio Watermarking & Speech Tech

Aalto University; Finland; Hybrid; EUR 43,000 - 51,000; Full time

Job summary

Aalto University invites applications for a Postdoctoral Researcher in Speech and Language Technology at the DICE department. The project COWAMA focuses on content‑based watermarking for AI‑generated audio, spanning speech, music, and audio signals.

You will work under Assistant Professor Lauri Juvela, collaborating with international researchers and contributing to open‑source software and reproducible research.

Qualifications

• We expect a doctoral-level researcher with expertise in speech or audio processing, ML, signal processing, CS or related field.
• Strong experience with deep learning and generative models.
• Programming skills suitable for modern ML research, preferably Python and DL frameworks.
• Interest in speech synthesis, music generation, neural audio codecs, diffusion models, or audio representation learning.
• Ability to conduct independent research and publish in peer‑reviewed venues.
• Commitment to open science and reproducible research with open-source software.

Responsibilities

• Develop content‑based audio watermarking methods for generated speech, music, and audio.
• Design detection and perceptual quality measures for perceptual fidelity.
• Explore audio content detection, music information retrieval, audio fingerprinting, and self‑supervised representations.
• Integrate watermarking into deep audio generative models and diffusion/audio language models.
• Assess embedding of watermark information in long‑term content like semantics and prosody.
• Develop robustness evaluation methods and realistic attack scenarios for watermarks.
• Publish in leading venues and contribute to open-source software and models.

Skills

Deep learning

Generative models

Python

Communication skills

English fluency

Education

Doctoral degree in speech or audio processing, ML, signal processing, CS or related field

Tools

Python

Deep learning frameworks

Job description

Aalto University invites applications for a Postdoctoral Researcher in Speech and Language Technology at the DICE department. The project COWAMA focuses on content‑based watermarking for AI‑generated audio, spanning speech, music, and audio signals.

You will work under Assistant Professor Lauri Juvela, collaborating with international researchers and contributing to open‑source software and reproducible research.