Yandex Vision

Machine-learning models for image analysis.
With Yandex Vision, you can create searchable e-document archives, automate your business workflow, or solve other tasks involving image analysis.
Access the service using simple REST and gRPC APIs.
  • Text recognition in images
    Yandex Vision recognizes text in scanned document pages and returns text blocks, lines, and separate words with their position in the image. You can also use it to detect text in photos, like road signs or license plate numbers.
  • More than 40 languages
    Yandex Vision can detect and recognize text in more than 40 languages, including Russian, English, German, Japanese, and many others.
    Image moderation
    Image classification technology helps you analyze the content your users upload. Use moderation properties to determine copyright infringement by the presence of the watermarks or to filter other inappropriate content.
    Face detection
    The service allows you to determine the presence of faces in the image and their number, as well as get their coordinates for further processing. For example, you can find all the faces in the photos uploaded by the user and suggest tagging people they know.
  • High-speed image processing
    Average processing time is 1-2 seconds per page. Each request can send multiple images containing text in one or more languages. This makes it faster to recognize text in different pages of a document or process a set of related photos.
  • Easy integration
    Yandex Vision can be easily integrated with your application using REST or gRPC APIs. You don’t need to have a background in machine learning.

Use cases

  • Many companies face the challenge of searching an archive of scanned documents that only exist in image format. Yandex Vision helps you recognize text in archived documents for indexing and searching.

  • Developers can use Yandex Vision and Yandex Translate to create applications with automatic text recognition and translation features for images.

  • You can use Yandex Vision to automate your business processes: scan and recognize text in standard types of documents, automatically add them to your document workflow system, and classify them using keywords or information from specific document sections. Along with recognized text, the service also returns the position of each text fragment within the image.

  • If your site or application gives users the ability to upload photos and share them online, Yandex Visions image classification can help you check the uploaded files and identify unwanted content. You can assess whether the image quality meets your requirements, detect copyright infringement by the presence of the watermarks, or find and block adult content.

Try Yandex Vision:

Get startedAll services