Keynotes

Anton Nijholt: BCI for Games - Games for BCI

Abstract

We survey recent research views on non-traditional brain-computer interfaces. That is, interfaces that can process brain activity input, but that are designed for non-medical purposes, that are meant to be used by ‘healthy’ users, that process other user input as well, and that even allow input, voluntarily or involuntarily, from multiple users. Application domains are entertainment, games, arts and domotics. Control of applications can be made more robust by fusing brain activity information with other information, either explicitly provided by a user (such as commands) or extracted from the user by interpreting his or her behavior (movements, posture, gaze, facial expression, nonverbal speech) or sensing (neuro-)physiological characteristics that provide information about a user's affective state. Multi-brain BCI with input from multiple users has also become a paradigm in artistic and game applications. Artistic interactive BCI applications may require audience participation. In game environments brain activity of various players can be used to control or adapt the game. Both competition and collaboration in serious games, entertainment games and artistic installations require fusion of EEG measurements from different subjects.

Bio

Anton Nijholt received his PhD in computer science from the Vrije Universiteit in Amsterdam. He held positions at various universities in and outside the Netherlands before settling down at the University of Twente in the Netherlands. His main research interests are human-computer interaction with a focus on entertainment, affect, humor and brain-computer interfacing (BCI). He edited various books, most recently on playful interfaces and BCI. Nijholt, together with more than fifty PhD students he supervised, wrote hundreds of journal and conference papers on these topics and acted as program chair and general chair of many large international conferences on affective and entertainment computing and on multimodal interaction. Nijholt is editor in chief of the specialty section Human-Media Interaction of the Frontiers in ICT and the Frontiers in Psychology Journals. Recently he got a position as a 'global research fellow' at the Imagineering Institute in Iskandar, Malaysia, where he will include 'smell' and 'taste' in his investigations in multimodal interaction and humor. His most recent publications are on humorous interactions in smart environments.

Tom Schaul: General Intelligence and Games

Abstract

Two desiderata for general intelligence are performance and generality. The first requires acting to achieve goals or solve problems; the second asks for agents that are competent on a broad range of tasks, or at least can learn to become so -- with minimal teaching signal, if possible. Games offers both precisely measurable performance and huge diversity, in a controlled setting. Recent research on AI and games started to emphasize the aspect of generality, leading to breakthroughs on board games (e.g. AlphaGo) as well as video games (e.g. DQN on Atari) -- with general-purpose learning methods that minimize game-specific knowledge. I will talk about two of their key ingredients, namely reinforcement learning (RL) and deep neural networks, and discuss why they are so generally applicable, and how to combine them effectively. I’ll also discuss a major limitation, namely that they are very data-hungry -- another reason why games (with their infinite data) are such appropriate domains.

Bio

Tom Schaul is a senior researcher in reinforcement learning at Google DeepMind. He did his PhD with Jürgen Schmidhuber at IDSIA and his Postdoc with Yann LeCun at NYU. He has published many areas of AI, including deep learning, optimization algorithms, artificial curiosity, evolutionary algorithms, and most recently on deep and hierarchical RL. He thinks that substantial progress on general AI is possible, and that games are perfect benchmark domains for that.

Innes McKendrick: An iterative approach to the evaluation, application and design of PCG techniques

Abstract

Procedural content generation is a massively powerful tool to games development, enabling small teams to create virtual worlds on a scale that might otherwise require hundreds of developers. However, producing assets on-the-fly that meet the aesthetic, performance and gameplay needs of a modern title presents a significant challenge, requiring specialised knowledge not only from programmers but from artists, animators and designers as well.

In this talk I will discuss our iterative approach to the evaluation, application and design of procedural content techniques, with the goal of elevating the work of artists and directors, rather than replacing them. Drawing on examples from the development of sci-fi exploration game No Man's Sky, I will compare both specialised and general purpose methods of generating author-driven content which is varied, interesting and performant. Additionally, considering procedural generation in the context of game development , I will discuss the importance of workflow and automated testing in applying these techniques effectively.

Bio

Innes McKendrick is a programmer at UK indie games studio Hello Games, currently working on the universe-spanning exploration game No Man's Sky. His involvement with the title has spanned many areas, from gameplay to effects and visuals, however his major interest remains with procedural content generation.