AIbase
EN
Home
Detr Resnet 50 Fine Tuned Loc 2023
D

Detr Resnet 50 Fine Tuned Loc 2023

Developed by biglam
Object detection model fine-tuned on the loc_beyond_words dataset based on facebook/detr-resnet-50
Object Detection
Transformers
 OtherOpen Source License:Apache-2.0#Historical Newspaper Object Detection#DETR Fine-tuned Model#Document Layout Analysis
Downloads 24
Release Time : 4/13/2023

Model Overview

This model is an object detection model based on the DETR (Detection Transformer) architecture, specifically fine-tuned for object detection tasks in historical newspaper documents.

Model Features

Transformer-based Object Detection
Utilizes the DETR architecture, combining the strengths of Transformer and CNN for end-to-end object detection.
Historical Document Optimization
Specially fine-tuned for historical newspaper documents, suitable for cultural heritage digitization scenarios.
Efficient Training
Achieved a validation loss of 0.8784 after 100 training epochs, demonstrating good convergence.

Model Capabilities

Object detection in images
Historical document analysis
Newspaper layout element recognition

Use Cases

Cultural Heritage Digitization
Historical Newspaper Layout Analysis
Automatically identifies elements such as article areas, images, and advertisements in historical newspapers.
Document Processing
Document Structure Analysis
Identifies different content regions in complex documents.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase