This dataset encompasses a diverse range of cultural concepts from five different languages and cultural backgrounds: Indonesian, Swahili, Tamil, Turkish, and Chinese.
Specifically, it includes 236 concepts in Chinese, 128 in Indonesian, 202 in Swahili, 178 in Tamil, and 178 in Turkish, with each cultural concept represented by at least two images, totaling 2,235 high-quality images.
Each culture comprises ten primary categories, covering festivals, music, religion and beliefs, animals and… See the full description on the dataset page:
https://huggingface.co/datasets/zhili312/multimodal-cultural-concepts.
