Intern - Master Thesis - Machine Learning

Job Description

SEED is a pioneering group within Electronic Arts, combining creativity with applied research. We explore, build, and help determine the future of interactive entertainment.
Our mission is to bring and foster inspiring innovation for our games and our players. We do this through research, prototypes, exploration, and never-ending curiosity. We exist to educate and be great partners. We work with game teams and publish and present our research within our industry and to the public. We pursue long-term applied research and are impactful by presenting and delivering leading technology to game teams and technology groups inside EA.

Master Thesis - Machine Learning

We now offer a couple of Master Thesis opportunities in the area of Machine Learning for the next generation of games.

Together with our researcher from SEED, you will work on one of the following topics:
• Conversation Based Avatar Behavior: In this project, you will study how to animate realistic nonverbal behaviors for avatars in group conversations by training an ML-based generative model. One can sample, e.g., facial expression and body pose parameters, conditioned on features stemming from a conversation. Such features may include speech features from the dialogue or monologue, extracted/annotated emotions, personality parameters, and a processed script. A well-working system with these properties could have a range of use cases in-game animation, including simplifying workflows for scenes with group conversations and leading to increased immersion and realism.
• Automated LOD reduction and evaluation using Machine Learning: You will use Machine Learning (ML) to reduce LOD for games and assess the quality of autogenerated LODs.There are a few automated approaches for creating LODs, but these are not tailored for games and need to be assessed further in terms of quality. We need to be tailored to style and optimization. The hypothesis is that if enough training data is provided it should be possible to find ways to reduce LOD for 3D assets in game. This yields two research questions: (1) Automating the generating of reduced LOD, and (2) Evaluating generated LODs quality.

What We Look For

We are looking for students with a background in computer science and machine learning techniques and have an interest to work on state-of-art tools for games. Experience in Unity, Unreal, Maya, or Houdini will also improve your merits.

How To Apply

Please apply online with a CV and cover letter. Also, state what project you would like to work on and explain your interest based on your interests and skills.

This internship is unpaid and temporary employment. It needs to be associated with a university education or any other type of education that has an internship period planned into the education plan, for example, LiA. We are not able to provide relocation services for this role, so we will only consider candidates already residing in Sweden.

We look forward to receiving your application!

Jobcode: Reference SBJ-gx3181-3-140-185-123-42 in your application.