AIDOVECL
收藏资源简介:
AIDOVECL数据集由伊利诺伊大学厄巴纳-香槟分校创建,是一个AI生成的车辆图像数据集,旨在解决眼水平分类和定位问题。数据集包含超过15000张AI生成的车辆图像,这些图像通过检测和裁剪手动选择的种子图像生成,并使用高级外绘技术模拟真实世界条件。数据集的创建过程包括车辆检测、图像裁剪、外绘和质量评估,确保视觉保真度和上下文相关性。该数据集主要应用于自动驾驶、交通分析和城市规划领域,旨在提高机器学习模型在多样化操作场景下的分类和定位性能。
The AIDOVECL dataset, developed by the University of Illinois Urbana-Champaign, is an AI-generated vehicle image dataset tailored for eye-level classification and localization tasks. It contains over 15,000 AI-generated vehicle images, which are generated by detecting and cropping manually selected seed images, with advanced outpainting techniques applied to simulate real-world conditions. The dataset's creation pipeline encompasses vehicle detection, image cropping, outpainting and quality assessment, ensuring visual fidelity and contextual relevance. Primarily deployed in the fields of autonomous driving, traffic analysis and urban planning, this dataset is designed to enhance the classification and localization performance of machine learning models across diverse operational scenarios.




