Skip to content
View AbhiPandit1's full-sized avatar
  • Delhi

Block or report AbhiPandit1

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
AbhiPandit1/README.md

neofetch

Hugging Face LinkedIn LeetCode Email Portfolio


🔬 Research — in preparation

  • Reading the Unreadable — a benchmark for handwritten text recognition on unseen historical Danish hands, comparing trained specialists against general vision-language models (with Andy Stauder, Transkribus / READ-COOP).
  • Cross-lingual historical HTR (Danish ↔ Swedish) — era-specialist recognisers and a Danish↔Swedish transfer study on 16th–19th-century Nordic court and church records (with Uppsala University; target IJDAR).

🤗 Open models & data

Historical Danish HTR on the Hugging Face Hub — YOLO + kraken line segmenters, a 160k-line synthetic dataset, and a live demo. Segmentation is the hidden tax in historical HTR.


Pinned Loading

  1. ai-service-python ai-service-python Public

    AI proctoring microservice: YOLO-based CV detection (multi-face, look-away, phone/object) over webcam frames, served via FastAPI and Dockerized.

    Python 1

  2. danish-htr-synthetic danish-htr-synthetic Public

    160k synthetic 18th-century Danish handwriting lines with exact ground truth (HTR training data)

    1

  3. htr-united htr-united Public

    Forked from HTR-United/htr-united

    Ground Truth Resources for the HTR of patrimonial documents

    Python 1

  4. danish-htr-line-segmenter danish-htr-line-segmenter Public

    Runnable domain-adapted YOLOv8 line segmenter for historical Danish HTR (mAP50 ~0.79); loads weights from the Hugging Face Hub.

    Python 1