Lukaszl/pl-mixed-docs-ocr-dataset-100-v1-results
Viewer • Updated • 1.01k • 36
Polish OCR Benchmarks – Results
Collection of OCR benchmark result datasets focused on Polish-language and difficult real-world documents.
Note Evaluation results for OCR models on diverse Polish document types with mixed layouts, structural variation and real-world OCR challenges.
Lukaszl/pl-government-docs-mix-ocr-dataset-v1-results
Viewer • Updated • 1.02k • 46Note Benchmark results for OCR models evaluated on difficult Polish government and official-style documents with mixed layouts, forms, stamps and real-world scan artifacts.
Lukaszl/eu-law-ocr-dataset-pl-100-v1-results
Viewer • Updated • 1.01k • 20Note OCR benchmark results on difficult Polish-language EU legal and institutional documents with dense text and formal layouts.
Lukaszl/clearocr-benchmark-pl-insurance-terms-v1-results
Viewer • Updated • 11 • 56Note Benchmark results for OCR models evaluated on difficult Polish insurance and legal-style documents with dense text, small print and real-world readability challenges.
Lukaszl/pl-newspaper-pages-ocr-dataset-100-v1-results
Viewer • Updated • 1.01k • 39Note Benchmark results for OCR models evaluated on difficult Polish newspaper and magazine-style pages with multi-column layouts, headlines, embedded images and real-world editorial complexity.
Lukaszl/clearocr-benchmark-fiszki91-v1-results
Viewer • Updated • 918 • 44Note Benchmark results for OCR models evaluated on difficult Polish archival text snippets, featuring small print, clipping-style layouts, handwritten notes, underlines and scan artifacts.