hq_fer2013notest Open-source Image Classification Model - Free Deployment for Accurate Facial Expression Recognition

Hq Fer2013notest

Developed by Piro17

An image classification model based on ViT architecture, fine-tuned on the FER2013 dataset for facial expression recognition tasks.

Image Classification

Transformers

Open Source License:Apache-2.0 #ViT fine-tuned model #Image emotion recognition #High-accuracy classification

Downloads 37

Release Time : 2/18/2023

Model Overview

This model is a fine-tuned version of Google's ViT-base-patch16-224-in21k pre-trained model, specifically designed for facial expression recognition in image classification tasks.

Model Features

High accuracy

Achieves 70.52% accuracy on the FER2013 dataset

ViT-based architecture

Utilizes Vision Transformer architecture with powerful feature extraction capabilities

End-to-end training

Learns features directly from raw images without complex preprocessing

Model Capabilities

Facial expression recognition

Image classification

Emotion analysis

Use Cases

Affective computing

Facial expression recognition

Recognizes facial expressions of people in images

Can identify multiple basic expressions (e.g., happiness, sadness, anger, etc.)

Human-computer interaction

Emotional feedback system

Adjusts interaction methods based on user expressions

🚀 hq_fer2013notest

This model is a fine - tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset, achieving good results in image classification.

🚀 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: 0.8294
Accuracy: 0.7052
Precision: 0.7048
Recall: 0.7052
F1: 0.7036

📚 Documentation

Model Information

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

Model Index

Name: hq_fer2013notest
- Results:
  - Task:
    - Name: Image Classification
    - Type: image - classification
  - Dataset:
    - Name: imagefolder
    - Type: imagefolder
    - Config: default
    - Split: train
    - Args: default
  - Metrics:
    - Name: Accuracy, Type: accuracy, Value: 0.7052268506235075
    - Name: Precision, Type: precision, Value: 0.7048074435355876
    - Name: Recall, Type: recall, Value: 0.7052268506235075
    - Name: F1, Type: f1, Value: 0.7036260157126459

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
1.2982	1.0	353	1.2708	0.5635	0.5107	0.5635	0.5168
1.0218	2.0	706	1.0159	0.6411	0.6397	0.6411	0.6301
0.9437	3.0	1059	0.9452	0.6631	0.6698	0.6631	0.6556
0.8282	4.0	1412	0.8873	0.6829	0.6798	0.6829	0.6743
0.7717	5.0	1765	0.8612	0.6884	0.6888	0.6884	0.6835
0.7678	6.0	2118	0.8473	0.6985	0.6989	0.6985	0.6966
0.7096	7.0	2471	0.8363	0.7018	0.7001	0.7018	0.6989
0.6803	8.0	2824	0.8333	0.7036	0.7036	0.7036	0.7019
0.6521	9.0	3177	0.8309	0.7050	0.7039	0.7050	0.7028
0.6671	10.0	3530	0.8294	0.7052	0.7048	0.7052	0.7036

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.

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