The open-source sketch classification model beit - sketch - classifier - pt - metaset - 2 achieves an accuracy of 82.82% after fine-tuning!

Beit Sketch Classifier Pt Metaset 2

Developed by kmewhort

A sketch classification model based on Microsoft's BEiT architecture, fine-tuned on an unknown dataset with an accuracy of 82.82%

Image Classification

Transformers

Open Source License:Apache-2.0 #Sketch Classification #High Accuracy #Transfer Learning

Downloads 22

Release Time : 1/17/2023

Model Overview

This model is a fine-tuned version of microsoft/beit-base-patch16-224-pt22k-ft22k, primarily used for sketch classification tasks.

Model Features

High Accuracy

Achieves a classification accuracy of 82.82% on the evaluation set

Based on BEiT Architecture

Utilizes Microsoft's BEiT Vision Transformer architecture with powerful feature extraction capabilities

Efficient Fine-tuning

Fine-tuned on the base model to adapt to specific sketch classification tasks

Model Capabilities

Sketch Classification

Image Feature Extraction

Use Cases

Creative Design

Automatic Sketch Classification

Automatically classifies and organizes sketches drawn by designers

Accuracy: 82.82%

Education

Drawing Teaching Assistance

Identifies types of sketches drawn by students to provide targeted guidance

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.8028	1.0	76608	0.7586	0.8007
0.7168	2.0	153216	0.6983	0.8154
0.6357	3.0	229824	0.6676	0.8240
0.5707	4.0	306432	0.6606	0.8276
0.4254	5.0	383040	0.6703	0.8282

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご

Beit Sketch Classifier Pt Metaset 2

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 beit-sketch-classifier-pt-metaset-2

🚀 Quick Start

📚 Documentation

Model description

Intended uses & limitations

Training and evaluation data

🔧 Technical Details

Training procedure

Training hyperparameters

Training results

Framework versions

📄 License