# clovaai/CRAFT-pytorch

Official implementation of Character Region Awareness for Text Detection (CRAFT)

Repository: https://github.com/clovaai/CRAFT-pytorch
Canonical: https://ross.abutalabs.com/products/craft-pytorch
Language: Python
License: MIT
License Family: permissive
Topics: ocr, detection, text-detection, craft, curved-text, ocr-detection, pytorch, cvpr2019
Last push: 2024-07-16T04:32:12+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2647, "days_push": 778, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3398, forks 926 (observed 2026-08-28T04:08:02.677983+00:00)

## What it is
Official PyTorch implementation of CRAFT (Character Region Awareness for Text Detection), a scene text detector that localizes text by predicting character regions and character-link affinities. It ships pretrained models for general multilingual text and IC15, with optional link refinement for sentence-level detection.

## Use cases
- detect text regions in natural scene images
- find bounding boxes of text for an OCR pipeline
- detect curved or rotated text in images
- run pretrained scene text detection on a folder of images
- extract text areas from photos before recognition
- benchmark text detection on ICDAR datasets

## When to choose
- you need a proven scene text detector with pretrained weights
- your images contain curved, rotated, or multi-oriented text
- you want character-level awareness to build polygon or box detections
- you need a PyTorch-based detector that integrates with a recognition model

## When to avoid
- you need to train the model from scratch - training code is not released
- you need text recognition, not just detection
- you want a maintained pipeline with frequent updates
- you need a non-PyTorch or lightweight CPU-only deployment

## Facets
- artifact type: library
- maturity: maintenance
- function: ocr, computer-vision, image-processing, deep-learning
- domain: computer-vision, image-processing
- platform: python, cross-platform
- tags: text-detection, scene-text, pytorch, cvpr2019, curved-text, deep-learning, ocr, gpu

## Member repositories
- clovaai/CRAFT-pytorch (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:02.677983+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:38:47.230892+00:00, confidence not recorded.
  - readme: https://github.com/clovaai/CRAFT-pytorch (fetched 2026-08-28T04:08:02.677983+00:00, sha 88c410dba286)
- Data as of 2026-08-30T08:39:29.467469+00:00.