Back to jobs
Job Description
- Improve quality of models for audio understanding and generation, including research on architectures, representations, training losses and paradigms, and test-time techniques for improved generation quality and efficiency.
- Unlock new audio capabilities in foundational models, both in pre-training and post-training data pipelines.
- Develop better evaluation methods (human evaluation, auto raters, automated metrics) to measure quality of open-ended audio tasks.
- Publish research at venues and contribute to Google DeepMind products.
- Collaborate across teams to advance research in sound understanding, joint audio-video generation, and audio editing.