# da03/Attention-OCR

Visual Attention based OCR

Repository: https://github.com/da03/Attention-OCR
Canonical: https://ross.abutalabs.com/products/attention-ocr
Language: Python
License: MIT
License Family: permissive
Last push: 2018-11-08T03:45:21+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3738, "days_push": 2855, "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 1117, forks 355 (observed 2026-08-28T04:03:38.880316+00:00)

## What it is
A Python/TensorFlow implementation of a visual attention-based OCR model that combines a sliding CNN, an LSTM, and an attention decoder to recognize text in images. It supports training on custom image-text datasets and includes a toy example based on Synth 90k.

## Use cases
- recognize text in scene images with ocr
- train a custom ocr model on my own image dataset
- read text from images using deep learning
- scene text recognition with attention model
- extract characters from cropped word images
- build a cnn lstm attention ocr model

## When to choose
- you need a trainable attention-based OCR model and can work with older TensorFlow versions
- you want to train text recognition on your own labeled image-text pairs
- you are researching CNN-LSTM-attention architectures for scene text

## When to avoid
- you need a maintained library or modern TensorFlow/PyTorch support
- you want a ready-to-use OCR tool for documents without training
- your project requires recent releases or active community support

## Facets
- artifact type: library
- maturity: abandoned
- function: ocr, machine-learning, deep-learning, image-processing
- domain: computer-vision, machine-learning
- platform: python
- tags: attention-model, cnn-lstm, scene-text-recognition, tensorflow, natural-language-processing, linux, macos

## Member repositories
- da03/Attention-OCR (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:38.880316+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 2026-08-30T06:41:53.081514+00:00, confidence not recorded.
  - readme: https://github.com/da03/Attention-OCR (fetched 2026-08-28T04:03:38.880316+00:00, sha 96de8dcd0293)
- Data as of 2026-08-30T08:39:29.467469+00:00.