metrika-labs/ru-ocr-benchmark-hard download history
metrika-labs/ru-ocr-benchmark-hard is an image to text dataset on the Hugging Face Hub. In the last 30 days it was downloaded 3,266 times (480 in the last 7 days), and 4,885 times in total. It ranks #11,308 among datasets by monthly downloads.
A benchmark for evaluating OCR in Russian, consisting of two parts: pages and crops. The pages part contains 155 images covering 31 source documents in five variants: originals, scans, low-quality images, print defects, and distortions. Reference text and coordinates for each character enable evalua
Open metrika-labs/ru-ocr-benchmark-hard on Hugging Face
Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.