issai/Central_Asian_Food_Dataset

中亚食品数据集（CAFD）是一个包含16,499个食品项目图像的新数据集，涵盖42个类别。该数据集旨在解决中亚地区食品图像识别公共资源的缺乏问题。数据集的类分布不平衡。已经在多个预训练模型上对数据集进行了训练，并对其在更大的食品数据集Food1K上的性能进行了评估。提供的预训练模型权重可供下载。该数据集和预训练模型对于开发针对中亚菜系的食物图像识别应用和个性化饮食干预系统非常有用。

The Central Asian Food Dataset (CAFD) is a new dataset containing 16,499 images of food items across 42 classes. It aims to address the lack of publicly available resources for food image recognition in Central Asia. The dataset is imbalanced with respect to class distribution. Several pre-trained models have been trained on this dataset and their performance has been evaluated against a larger food dataset (Food1K). The pre-trained model weights are available for download. The dataset and pre-trained models are useful for developing food image recognition applications specific to Central Asian cuisine and personalized dietary intervention systems.