Computer Vision Annotation Tool

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
Computer Vision Annotation Tool
CVAT version 0.4.0
CVAT version 0.4.0
Initial releaseJune 29, 2018; 13 months ago (2018-06-29)
Written inJavaScript, CSS, Python, HTML, Django
Operating systemWindows 7 or later, OS X 10.11 or later, Linux
Available inEnglish (US)
TypeImage and video annotation tool
LicenseMIT License[1]

Computer Vision Annotation Tool (CVAT) is a free, open source, web-based image and video annotation tool which is used for labeling data for computer vision algorithms. CVAT was developed for use by a professional data annotation team, with a user interface optimized for computer vision annotation tasks.[2]

CVAT supports the primary tasks of supervised machine learning: object detection, image classification, and image segmentation. CVAT allows users to annotate data for each of these cases.[3]

CVAT has many powerful features, including interpolation of bounding boxes between key frames, semi-automatic annotation using Deep Learning models both in Intel OpenVINO and TensorFlow Object Detection API formats, shortcuts for most critical actions, a dashboard with a list of annotation tasks, LDAP and basic access authentication, etc.[4]

CVAT is written mainly in JavaScript, HTML, CSS, Python, and Django. It is distributed under the MIT License, and its source code is available on GitHub[5].

See also[edit]


  1. ^ "LICENSE". Intel.
  2. ^ "Intel open-sources CVAT, a toolkit for data labeling". VentureBeat. 2019-03-05. Retrieved 2019-08-13.
  3. ^ "Computer Vision Annotation Tool: A Universal Approach to Data Annotation". 2019-03-04. Retrieved 2019-08-13.
  4. ^ User's guide of Computer Vision Annotation Tool, OpenCV, 2019-08-13, retrieved 2019-08-13
  5. ^ Powerful and efficient Computer Vision Annotation Tool (CVAT): opencv/cvat, OpenCV, 2019-08-13, retrieved 2019-08-13