BIRD1445丨A 200-category Fine-grained Bird Image Dataset

Solemnly declare: If you use this open source content in papers, books, academic reports and other works, please quote the following documents (the original link has the latest citation format):WANG Hongchang, XIAN Fengyu, XIE Zihui, DONG Miaomiao, JIAN Haifang. BIRD1445: Large-scale Multimodal Bird Dataset for Ecological Monitoring[J]. Journal of Electronics & Information Technology. doi: 10.11999/JEIT250647Authors: WANG Hongchang , XIAN Fengyu, XIE Zihui, DONG Miaomiao, JIAN HaifangAuthor' s unit: 1.Laboratory of Solid State Optoelectronics Information Technology, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China2.School of Physics and Electronics, Shandong Normal University, Ji’nan 250358, ChinaFunds: The National Science and Technology Major Project (2022ZD0116304)This dataset is a subset of the BIRD1445 dataset from the paper titled "BIRD1445: A Large-Scale Multimodal Bird Dataset for Ecological Monitoring." It consists of a fine-grained image dataset covering 200 bird species. The data is primarily sourced from two channels: first, field data collected via smart monitoring devices deployed in nature reserves (with relevant geographic information removed); and second, supplementary image data legally obtained from public online resources.  The dataset is constructed around 200 bird species. During data processing, intelligent noise cleaning techniques were applied to deduplicate, filter low-quality images, and remove abnormal samples. Additionally, a fine-grained annotation system was designed based on expert knowledge, with annotations including species names. The dataset comprises a total of 10,000 annotated images, making it a fine-grained bird image dataset tailored for ecological scenarios.  This dataset effectively supports model training and evaluation for fine-grained image classification tasks, providing a reliable data benchmark for the application of foundational artificial intelligence models in ecological conservation and scientific research.