Shypd.aiThe largest open index of AI toolsHandwritten-text-recognition-for-apache-mxnet is an Open Source tool that allows you to train neural network models for end-to-end full-page handwriting recognition. It utilizes the Apache MXNet deep learning framework on the IAM Dataset.
Trending
About
This repository provides the resources to train neural network models for performing end-to-end full-page handwriting recognition. It leverages the Apache MXNet deep learning framework and is designed to work with the IAM Dataset. The pipeline involves three key steps: detecting the handwritten area in a form, detecting lines of handwritten texts, and recognizing characters while applying a language model for error correction. Pre-trained models are available, and the repository includes Jupyter notebooks for each step, making it suitable for researchers and developers in the field of OCR. It also details the setup process, including dependencies like SCLITE for WER evaluation and hnswlib.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
