CLIP meets GamePhysics: Towards bug identification in gameplay videos using zero-shot transfer learning
This work addresses the problem of efficiently identifying bugs in gameplay videos for game testers and developers, though it is incremental as it applies an existing model to a new domain.
The paper tackles the challenge of parsing and mining large repositories of gameplay videos for bug identification by proposing a search method that uses English text queries to retrieve relevant videos based solely on video content, leveraging the zero-shot transfer capabilities of the CLIP model without requiring data labeling or training, and it shows promising results on a dataset of 26,954 videos from 1,873 games.
Gameplay videos contain rich information about how players interact with the game and how the game responds. Sharing gameplay videos on social media platforms, such as Reddit, has become a common practice for many players. Often, players will share gameplay videos that showcase video game bugs. Such gameplay videos are software artifacts that can be utilized for game testing, as they provide insight for bug analysis. Although large repositories of gameplay videos exist, parsing and mining them in an effective and structured fashion has still remained a big challenge. In this paper, we propose a search method that accepts any English text query as input to retrieve relevant videos from large repositories of gameplay videos. Our approach does not rely on any external information (such as video metadata); it works solely based on the content of the video. By leveraging the zero-shot transfer capabilities of the Contrastive Language-Image Pre-Training (CLIP) model, our approach does not require any data labeling or training. To evaluate our approach, we present the $\texttt{GamePhysics}$ dataset consisting of 26,954 videos from 1,873 games, that were collected from the GamePhysics section on the Reddit website. Our approach shows promising results in our extensive analysis of simple queries, compound queries, and bug queries, indicating that our approach is useful for object and event detection in gameplay videos. An example application of our approach is as a gameplay video search engine to aid in reproducing video game bugs. Please visit the following link for the code and the data: https://asgaardlab.github.io/CLIPxGamePhysics/