A wide variety of functions for you

ABEJA Platform Annotation is easy to operate and
brings outstanding results beyond your expectations.

  • Assistance by pre-trained models enhances annotation efficiency

    Assistance by pre-trained models rationalizes annotation work and helps to prevent errors. Also, it is a powerful function for reviewing labels.

    Contact us

    Assistance by pre-trained models enhances annotation efficiency
  • 10+ templates

    ABEJA Platform Annotation offers over 10 types of templates such as Segmentation, Classification, Voice, Video, Text. It reflects user insights.

    Contact us

    10+ templates
  • Smooth quality control process

    Your Project achieves high quality with "Three-layered control".

    Contact us

    Smooth quality control process

Abundant human resources for you

In need of training data but no human resource or time for annotation?
Please do not hesitate to contact ABEJA. Our experienced team will take care of you from data creation progress control to completion, no matter how big your project is.

Contact us

"Adding value to data" and beyond

ABEJA has been supporting clients with not only adding value to existing data but also generating new data by every measure through annotation.
Data goes through the hands of various people and gains value. It changes the society and harvests more value.

ABEJA wants to support such an ecosystem.
If you want to collect, enhance, and generate data, or to understand the value of the data, please do not hesitate to contact ABEJA.

Frequently asked questions
about ABEJA Platform Annotation.

Frequently asked questions about ABEJA Platform Annotation

  • What is "Annotator"?

    Annotators are specialists who check actual data and assign labels to it to create a labeled training dataset.

  • What types of annotation can be done?

    We have a wide variety of templates from image classification and object detection to video/text, audio, contour (semantic segmentation) to make various training datasets.
    We listen to our users and we plan to further increase the variety.

  • You say that it's possible to perform multiple annotation tasks simultaneously on cloud. What is the advantage?

    In comparison to general tools that need to be installed, it is more efficient and integral to work on cloud even with big projects. Moreover, more than one annotator can annotate one image data, so you can produce training datasets with even higher quality.

  • We want to use an outsourcing service but we don't even know what kind of annotation we need. Can we receive some support for that?

    Don't worry. We have consulting services as an option. We are happy to help you in various aspects such as finding the necessary annotation type, how to use training data, etc.

  • Can you make a quotation?

    Please feel free to request for a quotation. For outsourcing services, if possible, please provide us with some sample data as well and we will be able to respond to you more quickly.

  • Can you explain the process prior to the start of an outsourcing project?

    At first, we run a trial. We perform annotation using about 10 sample data extracted from an actual dataset. When the training data is ready, you can check its quality and we can also discuss about the labelling accuracy.

  • We want to train a model using the training dataset. What do we need to do?

    If you want to take the next step with the dataset such as to perform training, inference, develop inference API, etc., you can use our Deep Learning Platform, "ABEJA Platform". Demonstration of the platform is available. Please contact us for details.

  • We want to know about the contract details regarding the usage of the service.

    Information on our "Terms of Use" can be found here.

DownloadGet an e-book