Supported models

Yandex Vision OCR provides image classification based on a model that is trained on a specific set of images. The model is specified in the classificationConfig configuration in the request.

You can only include one of the two property types in the model:

• Mutually exclusive: The sum of probabilities of all model properties is equal to one.
• Independent: The probability of each property is equal to any value between zero and one.

Currently, the following models are supported:

• quality: Assesses the quality of an image.
• moderation: Estimates whether an image meets moderation requirements.

qualityquality

This model assesses the quality of an image.

With this model, you can:

• Filter low-quality images, for example, to exclude them from search results.
• Moderate images for a photo contest and select only high-quality ones.
• Compare similar images and choose the one with the best quality.

Try this model.

Type of properties: Mutually exclusive.

Properties:

• low: Low-quality image with clearly visible artifacts.
• medium: Regular quality image with minor artifacts.
• high: High-quality image without visible artifacts.

Examples of classification results:

"properties": [
    {
        "name": "low",
        "probability": 0.00039277845644392073
    },
    {
        "name": "high",
        "probability": 0.93234682083129883
    },
    {
        "name": "medium",
        "probability": 0.06726035475730896
    }
]

"properties": [
    {
        "name": "low",
        "probability": 0.072797946631908417
    },
    {
        "name": "high",
        "probability": 0.19245840609073639
    },
    {
        "name": "medium",
        "probability": 0.73474365472793579
    }
]

"properties": [
    {
        "name": "low",
        "probability": 0.7901121973991394
    },
    {
        "name": "high",
        "probability": 0.004683326929807663
    },
    {
        "name": "medium",
        "probability": 0.20520441234111786
    }
]

moderationmoderation

This model evaluates how well the image matches the moderation properties.

With this model, you can:

• Filter adult content.
• Detect illegal content.
• Demonstrate a copyright violation warning for images with watermarks.
• Check the image for text content so you can make use of text recognition.

Try this model.

Type of properties: Independent.

Properties:

• adult: Image contains adult content.
• gruesome: Image contains shocking content.
• text: Image contains text.
• watermarks: Image contains watermarks.

Example of a classification result:

"properties": [
    {
        "name": "adult",
        "probability": 0.0017486262368038297
    },
    {
        "name": "gruesome",
        "probability": 0.0002884353743866086
    },
    {
        "name": "text",
        "probability": 0.13789896667003632
    },
    {
        "name": "watermarks",
        "probability": 0.99908816814422607
    }
]