HQ_fer2013notestaugM Open Source Image Classification Model - Free Deployment for Precise Image Classification

Home

Hq Fer2013notestaugm

Developed by Piro17

A fine-tuned image classification model based on ViT architecture, excelling on the FER2013 dataset

Image Classification

Transformers

Open Source License:Apache-2.0 #Fine-tuned Vision Model #High-precision Classification #FER2013 Dataset

Downloads 17

Release Time : 2/19/2023

Model Overview

This model is a fine-tuned version of the google/vit-base-patch16-224-in21k pre-trained model for image classification tasks, primarily used for facial expression recognition

Model Features

High Accuracy

Achieves 69.98% accuracy on the FER2013 dataset

Based on ViT Architecture

Utilizes Vision Transformer architecture with powerful image feature extraction capabilities

Fine-tuned

Performance gradually improved through 10 training epochs

Model Capabilities

Image Classification

Facial Expression Recognition

Emotion Analysis

Use Cases

Affective Computing

Facial Expression Recognition

Recognizes facial expressions of individuals in images

Achieves 69.98% accuracy on the FER2013 dataset

Human-Computer Interaction

Emotional Feedback System

Used to detect user emotional states

🚀 hq_fer2013notestaugM

This model is a fine - tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It provides a high - performance solution for image classification tasks, leveraging the power of pre - trained models and fine - tuning techniques.

🚀 Quick Start

This model is a fine - tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 0.8297
Accuracy: 0.6998
Precision: 0.7022
Recall: 0.6998
F1: 0.6999

📚 Documentation

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 1e - 05
train_batch_size: 32
eval_batch_size: 32
seed: 17
gradient_accumulation_steps: 4
total_train_batch_size: 128
optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
1.2858	1.0	353	1.2814	0.5545	0.5432	0.5545	0.5122
1.0247	2.0	706	1.0343	0.6288	0.6235	0.6288	0.6136
0.9403	3.0	1059	0.9500	0.6607	0.6592	0.6607	0.6522
0.8501	4.0	1412	0.8971	0.6803	0.6761	0.6803	0.6760
0.8148	5.0	1765	0.8733	0.6857	0.6881	0.6857	0.6854
0.7898	6.0	2118	0.8526	0.6913	0.6911	0.6913	0.6888
0.7074	7.0	2471	0.8408	0.6959	0.6971	0.6959	0.6953
0.7273	8.0	2824	0.8361	0.6980	0.6971	0.6980	0.6949
0.6982	9.0	3177	0.8297	0.6998	0.7022	0.6998	0.6999
0.6994	10.0	3530	0.8287	0.6998	0.7002	0.6998	0.6991

Framework versions

Transformers 4.27.0.dev0
Pytorch 1.13.1+cu116
Datasets 2.9.0
Tokenizers 0.13.2

📄 License

This model is licensed under the Apache - 2.0 license.

Property	Details
Model Type	Fine - tuned version of google/vit-base-patch16-224-in21k
Training Data	imagefolder
Metrics	Accuracy, Precision, Recall, F1

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