Clash Royale Object Detection Dataset
收藏资源简介:
本研究设计了一个用于训练的生成物体检测数据集,针对Clash Royale游戏。数据集基于游戏场景的切片图像,通过AI辅助标注和分割技术生成。它包含不同层级的游戏单元,如地面部队、防御塔、空中部队等,以及它们的位置、类别、派别和健康信息。数据集的构建目的是为了训练物体检测模型,以支持非嵌入式AI在Clash Royale中的实时决策和游戏玩法。
This study develops a training-oriented generative object detection dataset tailored for the mobile game Clash Royale. The dataset is generated from sliced in-game scene images using AI-assisted annotation and segmentation technologies. It includes game units of various tiers, such as ground forces, defensive towers, air units, and more, alongside their positional coordinates, category labels, faction affiliations, and health status information. The dataset is constructed with the purpose of training object detection models to support non-embedded AI in implementing real-time decision-making and gameplay operations within Clash Royale.

- 1Playing Non-Embedded Card-Based Games with Reinforcement Learning西安交通大学人工智能与机器人研究所 · 2025年



