Midv720 Top 〈Easy〉

Unleashing the Apex of Anime Realism and Dynamic Composition.

MIDV-720 is a public dataset of 720 video clips of identity documents (IDs, passports, driver licenses, bank cards, etc.) captured under varying conditions to support research in document detection, recognition, and verification. It contains multiple document types, with frame-level annotations for document corners and ground-truth text for many fields. The dataset is widely used for training and evaluating OCR, document detection, and face cropping from ID photos. midv720 top