SDK for Computer Vision Annotation Tool (CVAT)

This package provides a Python client library for CVAT server. It can be useful for workflow automation and writing custom CVAT server clients.

The SDK API includes several layers:

• Server API wrappers (ApiClient). Located in at cvat_sdk.api_client.
• High-level tools (Core). Located at cvat_sdk.core.
• PyTorch adapter. Located at cvat_sdk.pytorch.
• Auto-annotation support. Located at cvat_sdk.auto_annotation.
• Miscellaneous utilities, grouped by topic. Located at cvat_sdk.attributes and cvat_sdk.masks.

Package documentation is available here.

Installation & Usage

To install a prebuilt package, run the following command in the terminal:

pip install cvat-sdk

To use the cvat_sdk.masks module, request the masks extra:

pip install "cvat-sdk[masks]"

To use the PyTorch adapter or the built-in PyTorch-based auto-annotation functions, request the pytorch extra:

pip install "cvat-sdk[pytorch]"

To install from the local directory, follow the developer guide.

After installation you can import the package:

import cvat_sdk