Kognitive Annotation Tool (KAT)

Kognitive Annotation Tool (KAT) is based on robust open source technologies and custom tools for annotating digital images and videos. KAT has been designed to provide users with an array of convenient annotation tools with powerful features.

KAT is a browser-based application for both individuals and teams that support different work scenarios. The main tasks of supervised machine learning can be divided into three groups:

Object detection
Image & video classification
Image & video segmentation

KAT allows you to annotate data for each of these cases. There are plenty of advantages of the tool.

Highlights

Web-based: Users don't need to install the app; they just have to open the tool's link in a browser if they want to create a task or annotate data.
Collaborative: Users can create a public task and split the work between other users.

Easy to deploy: KAT can be installed in the local network using Docker.
Automatic annotation: For example, users can employ interpolation between key frames.

Optional tools supported: TensorFlow Object Detection API.
TensorFlow Object Detection API, will be used to automatically annotate objects. Faster RCNN Inception ResNet V2 model trained on COCO (2-3 classes) is the default, but users can easily connect other models.

Solution Overview

Work Flow

Major Features

KAT is a cloud-based, on-premise, or containerized software solution that can be used to annotate production-grade training data for machine learning. It's designed to be used with specific types of data, such as image, video, text, audio, spreadsheet, or sensor data.

Dataset management
Annotation methods
Data quality control
Auto labeling services
Workforce management
Security
Integrated labeling services