Machine Learning Internship, Text-to-Spech Research
3 months ago(11/29/2018 1:28 AM)
AMZN Dev Cntr Poland sp. z.o.o
Location: Gdansk (Poland), Cambridge (UK)
The Alexa Speech team in Gdansk (Poland)/Cambridge (UK) develops new machine learning methods to improve naturalness and expressivity of speech synthesis. We collaborate closely with other science, engineering and product teams in Amazon to create new voices and speaking styles for Alexa and Polly, our AWS service for text-to-speech. Particular research areas include natural language processing, deep generative neural networks, variational auto-encoders and natural language generation.
We are recruiting curious and creative machine learning science interns who are looking to learn new skills and who are exited to collaborate with scientists and engineers to implement new scientific ideas.
The internship will involve experimenting with and implementing cutting-edge machine learning techniques to speech related technical challenges. Using state-of-the-art language models from NLP to capture a sophisticated representation of the meaning contained within text to improve expressivity of speech or generate richer text representations are especially relevant.
Current enrollment in a degree-granting college or university working towards a PhD in Machine Learning, Natural Language Processing, Applied Mathematics, Computer Science or a related field
At least one refereed academic publications in these areas
Good coding skills. Experience in Python is a plus
Good communication skills and the ability of working in a team
Ability to convey rigorous technical concepts and considerations to non-experts
Ability to distill problem definitions, models, and constraints from informal business requirements; and to deal with ambiguity and competing objectives