We have put together a ticketing API around a pretty useful machine learning model, called NSFW.

The model is capable of analyzing the image and providing reliable answers of the following questions:

• does the image contain Sexy stuff
• does the image contain Porn
• does the image contain Hentai
• does the image contain Drawing
• is the image Neutral

We have integrated this model into one of our microservices and made it available for public usage by providing 2 endpoints:

Create a prediction ticket:

Method: POST

Endpoint: https://ai.cybermall.dev/ml/image/NSFW

Payload:

{
  "src": "PATH_TO_IMAGE"
}

Response:

{
  "data": {
    "status": 1,
    "type": 1,
    "predictions": [],
    "_id": "TICKET_ID",
    "created": "2024-05-10T22:18:29.936Z",
    "modified": "2024-05-10T22:18:29.936Z"
  },
  "err": 0
}

Get your prediction results

Method: GET

Endpoint: https://ai.cybermall.dev/ml/predictions/TICKET_ID

Payload: none

Response:

{
  "data": {
    "_id": "663e66902cdbc8bd9c6aee29",
    "status": 3,
    "type": 1,
    "predictions": [
      {
        "className": "Sexy",
        "probability": 0.9057281017303467
      },
      {
        "className": "Porn",
        "probability": 0.08564005047082901
      },
      {
        "className": "Neutral",
        "probability": 0.007305879145860672
      },
      {
        "className": "Hentai",
        "probability": 0.0012525548227131367
      },
      {
        "className": "Drawing",
        "probability": 0.00007344265031861141
      }
    ],
    "created": "2024-05-10T18:25:20.761Z",
    "modified": "2024-05-10T18:25:20.761Z",
    "__v": 1
  },
  "err": 0
}

Response example

Please feel free to use this public service for your existing project.

We have selected the ticketing system over single request predictions so we could balance the load on this deployed model.

If you need professional consulting about how to build your own machine learning models to address your business needs or to make them available for the market, Give us a call