AIbase
EN
Home
Beit Base Patch16 224 Pt22k Ft22k Finetuned FER2013 5e 05
B

Beit Base Patch16 224 Pt22k Ft22k Finetuned FER2013 5e 05

Developed by lixiqi
Facial expression recognition model fine-tuned on the FER2013 dataset based on Microsoft's BEiT model
Image Classification
Transformers
Open Source License:Apache-2.0#Facial Expression Recognition#High-precision Classification#Sentiment Analysis
Downloads 18
Release Time : 1/9/2023

Model Overview

This model is a facial expression classifier based on the BEiT vision Transformer architecture, fine-tuned on the FER2013 dataset, capable of recognizing 7 basic facial expressions.

Model Features

Based on BEiT Architecture
Utilizes advanced vision Transformer architecture for efficient feature extraction through image patch processing
Facial Expression Recognition
Optimized specifically for facial expression recognition tasks, capable of identifying 7 basic emotions
Medium-sized Model
The base version achieves a good balance between computational efficiency and performance

Model Capabilities

Facial Expression Classification
Image Feature Extraction
Emotion Recognition

Use Cases

Human-Computer Interaction
Sentiment Analysis System
Used to analyze user facial expressions to assess emotional states
Achieved 68.33% accuracy on the FER2013 test set
Psychological Research
Emotional Response Study
Automatically analyzes changes in facial expressions of experiment participants
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