# MaybeShewill-CV/CRNN_Tensorflow

Convolutional Recurrent Neural Networks(CRNN) for Scene Text Recognition

Repository: https://github.com/MaybeShewill-CV/CRNN_Tensorflow
Canonical: https://ross.abutalabs.com/products/crnn_tensorflow
Language: Python
License: MIT
License Family: permissive
Topics: tensorflow, ctc-loss, ocr-recognition, sequence-recongnition, crnn-tensorflow, chinese-ocr
Last push: 2023-05-09T13:12:40+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": 3264, "days_push": 1212, "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 1039, forks 382 (observed 2026-08-28T04:03:19.922109+00:00)

## What it is
A TensorFlow implementation of CRNN (CNN + Bi-LSTM + CTC loss) for scene text recognition, based on the Shi et al. paper. It includes pretrained weights on Synth90k, evaluation scripts, and single-image inference tools.

## Use cases
- recognize text in natural scene images
- run OCR on cropped text line images
- train a CRNN model with CTC loss
- evaluate a text recognition model on Synth90k
- recognize Chinese text in images

## When to choose
- you need a classic CRNN text recognition implementation with pretrained Synth90k weights
- you want to train or fine-tune a CNN+LSTM+CTC recognizer in TensorFlow
- you need per-character and full-sequence accuracy metrics for OCR evaluation

## When to avoid
- you need modern transformer-based OCR or detection plus recognition end-to-end
- you require recent TensorFlow 2.x or PyTorch support
- you need actively maintained code with up-to-date dependencies

## Facets
- artifact type: library
- maturity: maintenance
- function: ocr, machine-learning, image-processing, deep-learning
- domain: computer-vision, image-processing, machine-learning, deep-learning
- platform: python
- tags: crnn, ctc-loss, scene-text-recognition, tensorflow, chinese-ocr, lstm, linux, gpu

## Member repositories
- MaybeShewill-CV/CRNN_Tensorflow (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:19.922109+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 2026-08-30T07:03:33.739675+00:00, confidence not recorded.
  - readme: https://github.com/MaybeShewill-CV/CRNN_Tensorflow (fetched 2026-08-28T04:03:19.922109+00:00, sha 5bf6847d347a)
- Data as of 2026-08-30T08:39:29.467469+00:00.