Midv-277 -

  • Preprocessing

  • Detection and rectification

  • OCR strategies

  • Face handling

  • Evaluation

  • Domain adaptation & robustness

  • Engineering practices

    • MIDV-277 is one of the datasets in the MIDV family (Mobile ID Document Video/Images) used as a public benchmark for identity-document analysis. It contains photographic images (and often associated annotations) of mock identity documents designed to support research on document detection, layout/type classification, text-field recognition (OCR), face detection/recognition, and related robustness tasks (capture conditions, rotations, occlusions, clutter).

    Key facts (concise):

    Note: If you need the exact file count, image resolution ranges, annotation schema or the current download link, tell me and I’ll fetch the precise, up‑to‑date details. MIDV-277

    This title is not for everyone. Avoid if you dislike:

    Watch for if you enjoy: