StanfordCars, FGVCAircraft, iNF200

本研究引入了三个新的细粒度数据集：StanfordCars、FGVCAircraft和iNF200，用于评估提出的方法在处理细粒度分类任务中的性能。StanfordCars包含16,185张汽车图像，FGVCAircraft包含10,000张飞机图像，而iNF200则专注于真菌类别的图像，包含200个不同的真菌类别。这些数据集的引入旨在解决现有模型在处理细粒度数据时面临的挑战，特别是在区分细微差异和特定细节方面的困难。通过这些数据集，研究展示了所提出的方法在提高模型对细粒度特征的识别能力方面的有效性，从而在相关应用领域如汽车识别、飞机型号识别和真菌分类中具有重要价值。

This study introduces three novel fine-grained datasets: StanfordCars, FGVCAircraft, and iNF200, which are employed to assess the performance of the proposed approach in fine-grained classification tasks. StanfordCars comprises 16,185 automobile images, FGVCAircraft contains 10,000 aircraft images, while iNF200 focuses on fungal-related images and encompasses 200 distinct fungal classes. The introduction of these datasets is designed to address the challenges faced by existing models when handling fine-grained data, particularly the difficulties in distinguishing subtle differences and specific details. Using these datasets, the study demonstrates the effectiveness of the proposed method in enhancing the model's ability to recognize fine-grained features, thereby holding significant value in relevant application domains such as vehicle recognition, aircraft type recognition, and fungal classification.