T

TF ID Large

Developed by yifeihu
TF-ID is a visual object detection model specifically designed for extracting tables and charts from academic papers, fine-tuned based on Florence-2
Downloads 9,893
Release Time : 7/10/2024

Model Overview

This model can identify tables and charts in academic papers and return their bounding box positions. The basic version simultaneously detects tables/charts and their caption texts

Model Features

High-precision Table/Chart Detection
Achieves 97.29% success rate on test sets, accurately identifying tables and charts in academic papers
Joint Caption Text Detection
The basic version can simultaneously detect tables/charts and their corresponding caption texts
Multiple Version Options
Offers various model variants including basic/large versions and with/without caption options

Model Capabilities

Academic paper image analysis
Table detection
Chart detection
Text region detection
Object detection

Use Cases

Academic Research
Paper Content Analysis
Automatically extracts table and chart information from papers for literature reviews and research analysis
Can accurately identify over 97% of tables and charts
Academic Knowledge Graph Construction
Serves as a preprocessing tool to help build academic knowledge bases containing visual elements from papers
Publishing Industry
Academic Journal Typesetting
Automatically detects visual elements in papers to assist journal typesetting work
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase