D

Devanagari PP OCRv3 Mobile Rec

Developed by PaddlePaddle
A dedicated text line recognition model for Devanagari in the PP-OCRv3_rec series developed by the PaddleOCR team, supporting Devanagari recognition with an average accuracy of 96.44%.
Downloads 168
Release Time : 6/6/2025

Model Overview

This model is a dedicated Devanagari model trained based on the PP-OCRv3 mobile recognition model, mainly used for recognizing Devanagari text lines.

Model Features

High accuracy
The average recognition accuracy reaches 96.44%, performing excellently in practical applications.
Lightweight
The model size is only 7.9M, suitable for mobile deployment.
Dedicated optimization
Specifically trained and optimized for Devanagari, resulting in better recognition effects.
Strict standards
If any character (including punctuation) in a line is incorrect, the entire line will be marked as an error to ensure high accuracy.

Model Capabilities

Devanagari text recognition
Text line detection
Multi-language OCR

Use Cases

Document processing
Digitization of Devanagari documents
Convert paper Devanagari documents into editable electronic text
Accuracy: 96.44%
Mobile applications
Devanagari recognition on mobile devices
Implement real-time Devanagari recognition on mobile devices
The lightweight model is suitable for mobile deployment
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase