B

Beit Base Patch16 224 Pt22k Ft22k Finetuned FER2013 8e 05

Developed by lixiqi
An image classification model based on the BEiT architecture, fine-tuned on the FER2013 dataset for facial expression recognition tasks
Downloads 87
Release Time : 1/9/2023

Model Overview

This model is a vision Transformer based on Microsoft's BEiT architecture, specifically fine-tuned for facial expression recognition tasks. It achieved an accuracy of 68.5% on the FER2013 dataset.

Model Features

Based on BEiT Architecture
Utilizes the BEiT (BERT pre-trained image Transformer) architecture, combining the advantages of vision Transformers
Facial Expression Recognition
Specifically optimized for facial expression recognition tasks
Medium-sized Model
Based on the BEiT-base architecture, balancing computational efficiency and performance

Model Capabilities

Image Classification
Facial Expression Recognition
Visual Feature Extraction

Use Cases

Affective Computing
Facial Expression Analysis
Analyze facial expressions in images to identify basic emotional states
Achieved 68.5% accuracy on the FER2013 dataset
Human-Computer Interaction
Emotion-aware Systems
Used to build interactive systems capable of sensing users' emotional states
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase