# bgshih/crnn

Convolutional Recurrent Neural Network (CRNN) for image-based sequence recognition.

Repository: https://github.com/bgshih/crnn
Canonical: https://ross.abutalabs.com/products/crnn
Language: Lua
License: MIT
License Family: permissive
Topics: torch7, sequence-recognition, computer-vision, machine-learning, ocr
Last push: 2019-03-04T05:12:37+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": 3910, "days_push": 2739, "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 2105, forks 547 (observed 2026-08-28T04:06:13.896103+00:00)

## What it is
An implementation of the Convolutional Recurrent Neural Network (CRNN), combining CNN, RNN, and CTC loss for image-based sequence recognition such as scene text recognition and OCR. It is built on Torch7 with Lua and includes pretrained models, a demo, and training tooling.

## Use cases
- recognize text in scene images
- build an OCR model for word images
- train a CRNN model on my own dataset
- lexicon-free text recognition from images
- implement CTC-based sequence recognition

## When to choose
- you need the reference CRNN implementation for scene text recognition
- you want to train a CTC-based text recognizer on custom data
- you are working in Torch7/Lua

## When to avoid
- you need a maintained PyTorch or TensorFlow OCR solution
- you require end-to-end text detection plus recognition
- your environment is not Linux with CUDA GPUs

## Facets
- artifact type: library
- maturity: maintenance
- function: ocr, machine-learning, deep-learning, image-processing
- domain: computer-vision, machine-learning, deep-learning
- platform: lua
- tags: torch7, ctc-loss, scene-text-recognition, sequence-recognition, cnn-rnn, natural-language-processing, linux, gpu, docker

## Member repositories
- bgshih/crnn (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:13.896103+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 2026-08-30T02:54:01.732392+00:00, confidence not recorded.
  - readme: https://github.com/bgshih/crnn (fetched 2026-08-28T04:06:13.896103+00:00, sha 3d9320da5b3d)
- Data as of 2026-08-30T08:39:29.467469+00:00.