作物病害领域多模态数据集(CDDM)
收藏资源简介:
CDDM数据集是由中国联通AI创新中心和中国联通数字技术公司创建的多模态数据集,包含137,000张不同作物病害的图片和100万条关于农业知识的问题-答案对。这些数据覆盖了从病害识别到管理实践的广泛农业知识。数据集的构建目的是为了推动农业研究领域的多模态学习技术发展,通过整合视觉和文本数据,促进精细化的问答系统开发,为农民和农业专业人士提供精确、有用的建议。
The CDDM dataset is a multimodal dataset developed by China Unicom AI Innovation Center and China Unicom Digital Technology Co., Ltd. It encompasses 137,000 images of various crop diseases and 1 million question-answer pairs concerning agricultural knowledge. These data cover a broad spectrum of agricultural knowledge ranging from disease identification to management practices. The dataset is constructed to advance the development of multimodal learning technologies in the field of agricultural research, and facilitate the development of fine-grained question-answering systems by integrating visual and textual data, thereby providing accurate and practical suggestions for farmers and agricultural professionals.

- 1A Multimodal Benchmark Dataset and Model for Crop Disease Diagnosis中国联通AI创新中心,中国联通数字技术公司 · 2025年



