Nsfw Image Detector

N

Nsfw Image Detector

Developed by LukeJacob2023

A NSFW (Not Safe For Work) content detection model fine-tuned based on Google Vision Transformer, capable of identifying 5 types of image content

Image Classification

Open Source License:Apache-2.0 #NSFW detection #High-precision classification #Content safety

Downloads 68.26k

Release Time : 1/22/2024

Model Overview

This model is used to detect Not Safe For Work (NSFW) content in images, classifying them into five types: drawings, hentai, neutral, porn, and sexy content.

Model Features

High-precision classification

Achieves 93.16% accuracy and 98.87% Top-K accuracy on the evaluation dataset

Multi-category recognition

Can distinguish between five types: drawings, hentai, neutral, porn, and sexy content

Optional resolutions

Provides 224 and 384 resolution versions to meet different scenario requirements

Model Capabilities

Image classification

NSFW content detection

Multi-category recognition

Use Cases

Content moderation

Social media content filtering

Automatically detects inappropriate content uploaded by users

Effectively reduces manual review workload

Home network filtering

Provides image classification support for parental control software

Helps block inappropriate content

Data preprocessing

Dataset cleaning

Automatically filters and classifies image datasets

Improves dataset quality

🚀 NSFW Image Detector

This NSFW Image Detector is a fine - tuned model based on Google's ViT architecture, which can accurately classify NSFW images and achieve high accuracy on the evaluation set.

🚀 Quick Start

This model is a fine - tuned version of [google/vit - base - patch16 - 224 - in21k](https://huggingface.co/google/vit - base - patch16 - 224 - in21k) on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 0.8138
Accuracy: 0.9316
Accuracy K: 0.9887

You can access 384 version on:

https://huggingface.co/LukeJacob2023/nsfw - image - detector - 384

✨ Features

High - accuracy Classification: Capable of accurately classifying NSFW images with high accuracy on the evaluation set.
Multiple Labels: Supports classification with labels ['drawings', 'hentai', 'neutral', 'porn','sexy'].

📚 Documentation

Model description

Labels: ['drawings', 'hentai', 'neutral', 'porn','sexy']

Intended uses & limitations

No detailed information provided in the original document.

Training and evaluation data

A custom dataset about 28k images, if you need to improve your domain's accurate, you can contribute the dataset to me.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 2e - 05
train_batch_size: 32
eval_batch_size: 32
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Accuracy K
0.7836	1.0	720	0.3188	0.9085	0.9891
0.2441	2.0	1440	0.2382	0.9257	0.9936
0.1412	3.0	2160	0.2334	0.9335	0.9932
0.0857	4.0	2880	0.2934	0.9347	0.9934
0.0569	5.0	3600	0.4500	0.9307	0.9927
0.0371	6.0	4320	0.5524	0.9357	0.9910
0.0232	7.0	5040	0.6691	0.9347	0.9913
0.02	8.0	5760	0.7408	0.9335	0.9917
0.0154	9.0	6480	0.8138	0.9316	0.9887

Framework versions

Transformers 4.36.2
Pytorch 2.0.0
Datasets 2.15.0
Tokenizers 0.15.0

📄 License

This project is licensed under the Apache - 2.0 license.

📦 Information Table

Property	Details
Base Model	google/vit - base - patch16 - 224 - in21k
Tags	generated_from_trainer
Datasets	imagefolder
Metrics	accuracy
Model Name	nsfw - image - detector
Task	Image Classification
Dataset Name	imagefolder
Dataset Type	imagefolder
Dataset Config	default
Dataset Split	train
Dataset Args	default
Accuracy	0.9315615772103526
License	Apache - 2.0

Featured Recommended AI Models

AIbase

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご

© 2025AIbase