Advanced Detection Models
SightCensor offers a suite of specialized machine learning models for content analysis, each designed for specific use cases and detection requirements.
4 Models
Image & video detection
0.5s Avg
Processing time per image
97.7%
Neutral detection accuracy
REST API
Easy integration
Nudity Classification
Binary classification model that determines the probability of an image containing sexually explicit content. Fast, efficient, and accurate.
Capabilities
- Multi-class classification: Neutral, Moderate, Anime, Pornography
- Confidence scores for each classification
- Optimal for high-volume screening
- Low false positive rate
- Frequent model updates and improvements
Best For: Initial content screening, categorisation, or identifying images that require further analysis.
Classification Results
Sexually Explicit Object Detection
Advanced model providing both image classification and precise object detection with bounding box coordinates for sexually explicit content.
Capabilities
- Binary classification (neutral & pornography)
- Genitalia detection: breasts, vagina, penis, bottom
- Bounding box coordinates for each detection
- Individual confidence scores per object
- Our most efficient detection model
Best For: Content moderation, or teams requiring granular insight into the nature of potentially explicit content.
Object Detection Results
Detailed Sexually Explicit Object Detection
Our most advanced model providing verbose object detection and precise localisation of explicit content with extended classification categories.
Capabilities
- Binary classification (neutral & pornography)
- Extended genitalia & body trait classification
- Detailed bounding box annotations
- Higher precision for forensic analysis
- Currently in BETA testing phase
Best For: Detailed forensic analysis, investigations, and situations requiring precise content location within images.
Detailed Analysis
- Higher precision for complex scenes
- More detailed forensic reporting
- Better handling of edge cases
Sexually Explicit Video Detection
Based on Model 3, our video detection model analyses video content frame-by-frame to detect and classify sexually explicit content within videos.
Capabilities
- Binary classification (neutral & pornography)
- Frame-by-frame analysis
- Genitalia detection with timestamps
- Confidence scores per frame
- Support for all common video formats
Best For: Video content moderation, forensic video analysis, and platform safety for video-sharing applications.
Video Analysis Report
Simple API Integration
Integrate SightCensor models into your application with our REST API
Example API Request
curl -X POST https://api.sightcensor.com/v1/analyze \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: multipart/form-data" \
-F "image=@/path/to/image.jpg" \
-F "model=model_2"Choose Your Detection Model
Start with Model 1 on our free plan, or unlock advanced detection with Professional and Enterprise tiers.