The BIOSCAN-1M Insect Dataset aims to expand the cataloging of insect biodiversity through a large dataset of one million hand-labeled insect images. This curated image dataset is primarily intended for training computer vision models to provide image-based taxonomic assessments. Each record is taxonomically classified by an expert and includes associated genetic information, such as raw nucleotide barcode sequences and assigned barcode index numbers. The ultimate goal is to create a comprehensive survey of global biodiversity.
