Life-long learning in game development processes

In the game industry, software projects extend over several years: for instance, a typical AAA game is developed for 3 to 5 years. To make the development process easier for the developers, tools are put at their disposal to help with, for instance, the artistic creation process or code integration.

Weakly Supervised Behavioral Modeling for Controllable AI Agents in Video Games

The project aims at developing a new type of Reinforcement Learning algorithm that would allow to retain more control over the artificial agent once its training is completed. This framework would combine modern unsupervised modeling techniques to capture the variability of a set of demonstrations and user-defined programmatic functions that can characterise particularly important factors of variations. Ultimately, the user would be able to specify to the learned agent what type of behavior should be executed at test-time.

Adversarial Reinforcement Learning for Scaling Synthetic Trajectory Imitation in Physics-based Character Animation

With techniques such as Motion Matching and large motion capture data sets, video game characters have become very realistic for user-controlled motion. The next stage towards realistic video game characters is to incorporate physical interaction into the characters behaviour so that it reacts to new environments in a realistic way.

Data-Driven and Anthropometric Local Editing of Facial Polygon Meshes

Editing the faces of 3D avatars is a difficult and important task. We will develop an approach that enables users

to perform local edits of faces through means of adjusting the values of anthropometric measurements. Such

measurements are derived from well-established research about the shape and proportion of faces. Based on a

data set of 3D scans of faces, our approach will understand trends among the measurements and the shape of

faces. With our approach, editing the face will be easier and more predictable.