T

Tablecell Htr

Developed by Kansallisarkisto
This model is designed to recognize handwritten text from text line images in table cells, particularly suitable for handwritten text recognition in Finnish death records and census records from the 1930s.
Downloads 39
Release Time : 9/12/2024

Model Overview

Fine-tuned from the Finnish National Archives' multi-century handwritten text recognition model and Microsoft's TrOCR model, specifically designed for recognizing handwritten text in specific types of table cells.

Model Features

Domain-Specific Optimization
Specially optimized for handwritten text in historical document table cells, with high recognition accuracy.
Based on TrOCR Architecture
Leverages the powerful capabilities of Microsoft's TrOCR model, fine-tuned with domain-specific data.
GPU Acceleration Support
Supports GPU-accelerated inference for improved processing speed.

Model Capabilities

Handwritten Text Recognition
Table Cell Content Extraction
Historical Document Digitization

Use Cases

Archive Digitization
Historical Population Record Transcription
Convert handwritten content from 1930s Finnish census records into searchable text.
Verified CER of 0.107, WER of 0.237
Death Record Digitization
Automatically recognize and transcribe handwritten information in historical death record tables.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase