They say never judge a book by its cover. Now, a group of researchers have taught an artificial intelligence (AI) system to just that but in case of video games.
Researchers Yuhang Jiang and Lukun Zheng recently published pre-print research (via TNW) paper titled “Deep learning for video game genre classification” wherein they talk about the creation of a large database and then using it to train an AI system to classify a video game by its cover.
Researchers say that they compiled a large dataset consisting of 50,000 video games from 21 genres made of cover images, description text, title text and genre information. They also evaluated the mage-based and text-based models for the task of genre classification for video games.
“Third, we developed an efficient and salable multi-modal framework based on both images and texts. Fourth, a thorough analysis of the experimental results is given and future works to improve the performance is suggested,” the researchers wrote in their paper.
Upon investigating, the researchers determined that the text-based models perform better in general than image-based models. They also determined that the multi-modal model consisting of both text-based and image-based models outperformed image-based or text-based models.