# High-precision OCR

An ultra-lightweight English text line recognition model developed by the PaddleOCR team, supporting the recognition of English and numeric characters

SLANeXt_wired is a deep learning model for table structure recognition, which can convert non - editable table images into editable table formats (such as HTML).

PP-OCRv5_server_det is the latest generation of text detection model developed by the PaddleOCR team. It is designed for high-performance application scenarios and supports the detection of text in various scenarios, including handwritten, vertical, rotated, and curved text. It can recognize multiple languages.

Llama-3.1-Nemotron-Nano-VL-8B-V1 is an advanced document intelligent vision-language model that can query and summarize images and videos, and supports multi-environment deployment.

Sapnous-6B is an advanced vision-language model that enhances perception and understanding of the world through powerful multimodal capabilities.

Aya Vision 32B is an open-weight 32B parameter multimodal model developed by Cohere Labs, supporting vision-language tasks in 23 languages.

Typhoon2-Vision is a Thai-supported visual language model capable of processing image and video inputs, specifically optimized for image-based applications.

PaliGemma 2 is an upgraded vision-language model developed by Google, combining the capabilities of Gemma 2, supporting image and text inputs to generate text outputs, suitable for various vision-language tasks.

TF-ID is a series of object detection models specifically designed to extract tables and figures along with their caption texts from academic papers.

TF-ID is a visual object detection model specifically designed for extracting tables and charts from academic papers, fine-tuned based on Florence-2

Pix2Text's Mathematical Formula Recognition (MFR) model, trained based on the TrOCR architecture, can convert mathematical formula images into LaTeX text representations.

Pix2Text's Mathematical Formula Detection (MFD) model for recognizing mathematical formulas in images

Sparrow is a document data extraction model fine-tuned on invoice data based on the Donut ML foundation model, designed to validate Donut's performance on enterprise documents.

Sparrow is a document data extraction tool fine-tuned on invoice data based on the Donut ML foundation model, designed to validate Donut's performance on enterprise documents.

This model is an image-to-text model based on the Apache-2.0 license, capable of converting image content into textual descriptions.

A document understanding model fine-tuned based on microsoft/layoutlmv3-base, excelling in document layout analysis tasks

MiniCPM-V 2.0 is a powerful multimodal large language model designed for efficient terminal deployment, built upon SigLip-400M and MiniCPM-2.4B and connected via a perceptual resampler.

A vision model for recognizing vehicle license plates, capable of extracting plate numbers from images.

Pix2Text's Mathematical Formula Recognition (MFR) model, trained based on the TrOCR architecture, capable of converting mathematical formula images into LaTeX text representations.

An OCR model fine-tuned based on microsoft/trocr-base-printed, specifically designed for recognizing license plates and timestamp information

A fine-tuned mathematical formula recognition model based on Nougat-small, excelling in extracting LaTeX formula code from images

CORD-v2 is a model for image-to-text tasks, primarily used for extracting and recognizing text content from images.

This model is outdated. It is recommended to use the official Nougat model. Nougat is an advanced vision-language model focused on document understanding and analysis.

An OCR model specifically designed for transcribing e13b MICR codes, fine-tuned based on Microsoft's TrOCR-large-stage1.

This is an image-to-text model based on the MIT license, capable of converting image content into descriptive text.

This is an image-to-text model based on the Apache-2.0 license, capable of converting image content into textual descriptions.

A model fine-tuned based on naver-clova-ix/donut-base, specific uses and functions require more information

A document layout analysis model fine-tuned based on the LayoutLMv3 architecture, specifically designed for document element classification tasks in the DocLayNet dataset.

Sparrow is a document data extraction model fine-tuned on invoice data based on the Donut ML foundation model, aimed at validating Donut's performance on enterprise documents.

This is an image-to-text model based on the MIT license, capable of converting image content into textual descriptions.

This is a Donut model fine-tuned on the CORD-v2 dataset, designed for image-to-text tasks, achieving an average accuracy of 0.901.

A document understanding model fine-tuned on the nielsr/funsd-layoutlmv3 dataset based on microsoft/layoutlmv3-base

A model fine-tuned based on naver-clova-ix/donut-base, specific purpose not explicitly stated

An invoice recognition model fine-tuned based on LayoutLMv3-base, trained on the wild_receipt dataset, excelling in extracting structured information from invoices.

TrOCR is a Transformer-based optical character recognition model designed for single-line text images, fine-tuned on multiple standard datasets.

A fine-tuned invoice information extraction model based on LayoutLMv3-base on the SROIE dataset, excelling in token classification tasks

A version fine-tuned on the WildReceipt dataset based on the LayoutLMv3-base model, designed for receipt key information extraction tasks

An invoice information extraction model fine-tuned based on the LayoutLMv3 architecture, demonstrating outstanding performance on the SROIE dataset

A document understanding model fine-tuned on the SROIE dataset based on Microsoft's LayoutLMv3-base model, excelling in extracting structured information from scanned documents

A version of LayoutLMv3-base fine-tuned on an invoice dataset for invoice information extraction