Re•Work 2021: Augmenting Automated Game Testing with Deep Reinforcement Learning
Machine Learning
SEED’s own Linus Gisslén presented this talk at Re•Work 2021 on January 29, 2021, during the Deep Learning 2.0 Virtual Summit.
Game testing is slow and expensive, and is becoming even more so as games grow in size and complexity. Typical approaches such as scripted bots are effective for some tasks but aren’t dynamic enough to fully test modern AAA games.
SEED is looking into using ML to extend the scope of game testing. In this talk, Linus describes our efforts to use deep reinforcement learning to improve automated testing of games.
Linus Gisslén is a senior research engineer in machine learning at SEED. His current research focus is on reinforcement learning (RL) and procedurally-generated content (PCG).
Check out all the events and find out more about Re•Work 2021.
Watch the video of the presentation below.
Download the Presentation Slides
Download the presentation slides as PDF (6 MB).