Car-1000
收藏资源简介:
Car-1000是一个大规模综合细粒度汽车分类数据集,由北京航空航天大学宇航学院、上海人工智能实验室和教育部新一代人工智能技术及其交叉应用 Key Laboratory (东南大学)共同创建。该数据集包含165个制造商的1000种不同车型,共有140312张图片,涵盖了从1960年代到2020年代的车型,具有广泛的时间覆盖范围。数据集采用三级层次标签系统,提供车型的属性和尺寸信息。Car-1000旨在为细粒度汽车分类研究提供一个新的视角和基准数据集。
Car-1000 is a large-scale comprehensive fine-grained vehicle classification dataset co-developed by the School of Astronautics of Beihang University, Shanghai AI Laboratory, and the Key Laboratory of New Generation Artificial Intelligence Technology and Its Cross-Disciplinary Applications under the Ministry of Education (Southeast University). This dataset includes 1000 distinct vehicle models from 165 manufacturers, with a total of 140,312 images, covering vehicle models ranging from the 1960s to the 2020s, boasting a wide temporal coverage. The dataset adopts a three-level hierarchical labeling system, providing attribute and size information for each vehicle model. Car-1000 aims to offer a novel perspective and benchmark dataset for fine-grained vehicle classification research.
- 1Car-1000: A New Large Scale Fine-Grained Visual Categorization Dataset教育部新一代人工智能技术及其交叉应用 Key Laboratory (东南大学), 北京航空航天大学宇航学院, 上海人工智能实验室 · 2025年



