AIbase
EN
Home
Beit Base Patch16 224 Pt22k Ft22k Finetuned FER2013
B

Beit Base Patch16 224 Pt22k Ft22k Finetuned FER2013

Developed by Celal11
A vision Transformer model based on BEiT architecture, fine-tuned on the FER2013 dataset for facial expression recognition tasks
Image Classification
Transformers
Open Source License:Apache-2.0#Micro-expression recognition#High-accuracy classification#Image emotion analysis
Downloads 19
Release Time : 1/8/2023

Model Overview

This model is a vision Transformer based on the BEiT architecture, specifically fine-tuned for facial expression recognition tasks, achieving 77.43% accuracy on the FER2013 dataset

Model Features

High-precision expression recognition
Achieves 77.43% accuracy on the FER2013 dataset
Based on BEiT architecture
Utilizes the BEiT vision Transformer architecture with powerful feature extraction capabilities
Pre-training + fine-tuning strategy
Pre-trained on the pt22k dataset, fine-tuned on the ft22k dataset, and specifically optimized on FER2013

Model Capabilities

Facial expression recognition
Image classification
Emotion analysis

Use Cases

Human-computer interaction
Emotion recognition system
Used to identify users' facial expressions to determine emotional states
77.43% recognition accuracy
Psychological research
Emotional response analysis
Used for recording and analyzing subjects' emotional responses in psychological experiments
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase