AIbase
EN
Home
Pedestrian Gender Recognition
P

Pedestrian Gender Recognition

Developed by NTQAI
This model is an image classification model fine-tuned on the PETA dataset based on the BEiT architecture, used for recognizing pedestrian gender with an accuracy of 91.07%.
Image Classification
Transformers
Open Source License:Apache-2.0#Pedestrian Attribute Recognition#High-precision Classification#Surveillance Scenarios
Downloads 93.78k
Release Time : 1/6/2023

Model Overview

A pedestrian gender classifier fine-tuned from Microsoft's BEiT-base model, specifically designed for long-distance pedestrian attribute recognition scenarios.

Model Features

High-precision Recognition
Achieves 91.07% accuracy on the PETA evaluation set with a loss of only 0.2170
Transfer Learning Optimization
Fine-tuned based on the pre-trained BEiT visual transformer model
Long-distance Adaptability
Specifically optimized for long-distance pedestrian attribute recognition scenarios

Model Capabilities

Pedestrian Gender Classification
Static Image Analysis
Attribute Recognition

Use Cases

Intelligent Surveillance
Mall Customer Flow Analysis
Statistics on the proportion of customers by gender
91.07% accuracy
Urban Security
Key Personnel Screening
Rapid screening based on gender characteristics
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