NSFW Image Detection
Automatically identify sexually explicit or sensitive content in images using AI models.
The NSFW Image Detection block is a powerful moderation tool that uses deep learning to scan images for Not Safe For Work (NSFW) content. It can identify nudity and specific exposed body parts, allowing you to automatically flag, hide, or delete sensitive media uploaded by users.
Core Concept
Think of this block as a digital bouncer. It checks every image at the door and reports back on how "revealing" the content is, so you can decide whether to let it into your application or block it.
Configuration
| Field | Required | Description |
|---|---|---|
| Input Image (Base64) | ✅ | The raw Base64 string of the image you want to scan for sensitive content. |
| Output Variable | ✅ | The name of the variable where the detection result object will be stored. |
This block requires a Premium subscription due to the high computational power needed for AI image analysis.
Error Handling
FIELD_MISSING— Triggered when no image data is provided for detection.PYTHON_ERROR— Triggered if the AI detection engine fails to initialize or process the image.PARSE_ERROR— Triggered if the results from the detection engine are unreadable.
Example
- Input Image:
{user_avatar_base64} - Output Variable:
moderation_result
Result: The variable {moderation_result} will contain an object like this:
{
"success": true,
"is_likely_nude": 0.89,
"detections": ["FEMALE_BREAST_EXPOSED", "BELLY_EXPOSED"]
}Recommendation for Moderation
To implement a balanced moderation system, we recommend checking if is_likely_nude is greater than 0.60:
- Strict: > 0.80
- Balanced: > 0.60
- Lenient: > 0.40