Automated visual bug detection

Visual regression testing automatically analyzes gameplay sessions to detect visual bugs (also known as glitches in a game context). Visual bugs are bugs that can be detected by just looking at them, such as those related to texture and lighting, but also more complex ones like those related to the physics engine (such as a flying horse) that require common-sense reasoning. The impact of visual bugs ranges from a minor annoyance to rendering a game unplayable. Visual bugs can be easy to spot for a human expert but difficult to detect for a computer. As a result, game development companies have to spend a considerable amount of resources on manual testing, which is tedious, expensive and prone to overlooking bugs. Having automated techniques for detecting visual bugs in video games would reduce the cost of testing drastically. In this project, we propose to use computer vision techniques and foundation models to automatically detect and highlight visual bugs in Ubisoft games in (1) screenshots from gameplay sessions and (2) real-time while playing the game.

Faculty Supervisor:

Cor-Paul Bezemer

Student:

Partner:

Ubisoft Divertissement

Discipline:

Computer science

Sector:

Information and cultural industries; Manufacturing

University:

University of Alberta

Program: