vit-base-nsfw-detector-384

ViT Large NSFW Detector (384-Res)

This is a fine-tuned version of google/vit-large-patch16-384, optimized for NSFW (Not Safe For Work) content detection. It leverages the high-resolution (384x384) Vision Transformer architecture to capture fine-grained visual details, making it highly effective for content moderation workflows.

Model Details

• Architecture: Vision Transformer (ViT) Large
• Base Model: google/vit-large-patch16-384
• Resolution: 384x384 pixels
• Task: Image Classification (Binary: NSFW vs Safe)
• License: MIT

Intended Use

This model is designed for developers and platforms needing automated content safety systems.

• Content Moderation: Automatically filter user-generated images.
• Safety Compliance: Ensure uploaded content adheres to community guidelines.
• Dataset Filtering: Clean large datasets by removing inappropriate imagery.

🚀 Usage

1. Using Transformers (Python)

You can run inference using the Hugging Face transformers library.

from transformers import AutoImageProcessor, AutoModelForImageClassification
from PIL import Image
import torch

# Load model and processor
model_name = "theusamaaslam/vit-base-nsfw-detector-384"
processor = AutoImageProcessor.from_pretrained(model_name)
model = AutoModelForImageClassification.from_pretrained(model_name)

# Load your image
image = Image.open("path/to/your/image.jpg")

# Preprocess
inputs = processor(images=image, return_tensors="pt")

# Inference
with torch.no_grad():
    outputs = model(**inputs)
    logits = outputs.logits
    predicted_class_idx = logits.argmax(-1).item()

# Get label
label = model.config.id2label[predicted_class_idx]
print(f"Prediction: {label}")

Usage for a distant image:

from transformers import AutoImageProcessor, AutoModelForImageClassification
from PIL import Image
import requests
import torch

# URL of the image to test
url = 'http://images.cocodataset.org/val2017/000000039769.jpg'

# Load image from URL
image = Image.open(requests.get(url, stream=True).raw).convert("RGB")

# Load Processor and Model from your Hub repo
repo_id = "theusamaaslam/vit-base-nsfw-detector-384"
processor = AutoImageProcessor.from_pretrained(repo_id)
model = AutoModelForImageClassification.from_pretrained(repo_id)

# Process image and inference
inputs = processor(images=image, return_tensors="pt")

with torch.no_grad():
    outputs = model(**inputs)

# Get probabilities and predicted label
logits = outputs.logits
probs = logits.softmax(dim=1)
predicted_class_idx = logits.argmax(-1).item()
predicted_label = model.config.id2label[predicted_class_idx]

print(f"Predicted class: {predicted_label}")
print(f"Probabilities: {probs}")

📬 Contact

Author: @theusamaaslam
Email: [[email protected]]
Organization: Nayatel (Pvt Ltd) AI & DevOps Team