INTRODUCTION
The Car-FG3K dataset is constructed by National Laboratory of Pattern Recognition, Institute of Automation, University of Chinese Academy of Sciences (NLPR, CASIA). This category is organized into a three-level hierarchy structure, incorporating makes, models and manufacture years. There are a total of 21,351 images covering 191 makes, 1,895 car models and 3,269 manufacture years. And we extract 10,676 images for training and 10,661 images for testing, covering 191 makes, 1,892 models and 3,255 manufacture years.
LICENSE
- The images and the corresponding annotation results can only be used for ACADEMIC PURPOSES. NO COMERCIAL USE is allowed.
- All images of the Car-FG3K database are obtained from the Internet which are not property of NLPR, CASIA. The NLPR is not responsible for the content nor the meaning of these images.
- Copyright © National Laboratory of Pattern Recognition, Institute of Automation, University of Chinese Academy of Sciences. All rights reserved.
All publications using Car-FG3K dataset should cite the paper below:
Hongyan Wu, Haiyun Guo, Qinghai Miao, Min Huang, and Jinqiao Wang. Graph Neural Networks Based Multi-Granularity Feature Representation Learning for Fine-Grained Visual Categorization. In MMM,2022.
DOWNLOAD
baidu drivepassword:7i0v