Artificial intelligence is a field of computer science that emphasizes the machine to learn and behave like a human. The machine learns the pattern from the historical data. The data may be of different types. For example audio, images, videos, texts, etc. We annotate the data or assign the Labels to train the machine learning models. There are many free open source annotation tools available on the Internet. Therefore, in this blog, I am providing a detailed overview of open-source annotation tools for images.

Artificial intelligence is a field of computer science that emphasizes the machine to learn and behave like a human. The machine learns the pattern from the historical data. The data may be of different types. For example audio, images, videos, texts, etc. We annotate the data or assign the Labels to train the machine learning models. There are many free open source annotation tools available on the Internet. Therefore, in this blog, I am providing a detailed overview of open-source annotation tools for images.

LabelImg

The most popular image annotation tool in the world. LabelImg has 15 thousand plus stars on the GitHub repository. Mainly, it saved the annotation into two formats i.e.; Pascal VOC and Yolo format. This is a user-friendly tool that has been written in Python and PyQt5 library for the graphical interface. This is the recommended tool for object bounding box annotation. You can check more details and tutorials of LabelImg at the following link.

Github Link: https://github.com/tzutalin/labelImg

Label Studio

If you want to use a single tool for the annotation of multiple types of data, then the label studio is best suited. The labeling tool helps to annotate the images, videos, audios, assign the labels on text, comparison between entities, and also you can use this tool for the time series data. This is one of the best tools which provides approximately all mediums to annotate. For the images, different use cases like classification, object detection, semantic segmentation, pose estimation tasks can be performed. The use cases for text tagging are classification and summarization. The tools also help in the audio tagging process for classification, emotion recognition, transcriptions, and the homogeneous segmentation of audio streams. The tool is very easy and rich in features. Label Studio is very popular on the GitHub platform with more than 3700 stars on this repository.

Website Link: https://labelstud.io/

Github Link: https://github.com/heartexlabs/label-studio

Makesense

Do you want an easy process to annotate the images without The complication of installation of tagging software on different operating systems? Then makesense.ai is a free online tagging website for you. This website provides a great automatic feature that helps the user during the labeling of the Images. You can load different artificial intelligence models which recommend marking to speed up the process. This website uses two AI models COCO SSD and Pose-Net. COCO SSD helps in marking rectangles on the object which is used in object detection. Pose-Net speeds up the process using automated points for pose estimation. This is the best deep learning-based project which also has a github repository with more than 1900 stars.

Website Link: https://www.makesense.ai/

Github Link: https://github.com/SkalskiP/make-sense

COCO Annotator

A web-based image annotation tool that is used for image segmentation, localization, and detection. The tool uses a special Coco format to export the annotations. COCO Annotator Tool is very rich in features. For example, It has semi-trained models to annotate the image automatically. It can generate a dataset from Google images. The tool allows the user to annotate the images using polygons, Rectangles, and free form curves. COCO annotator tool is one of the best open-source tools which is available on GitHub with more than 1300 stars. 

Github Link: https://github.com/jsbroks/coco-annotator

Computer Vision Annotation Tool (CVAT)

CVAT is a very famous annotation tool used for Computer vision tasks. The tool can perform annotations on images and videos. CVAT has very good documentation that explains how to use the CVAT annotation tool. The demo of the Computer Vision Annotation Tool (CVAT) is available online at cvat.org without local installation on any desktop. The automatic labeling option is also available which has been trained on deep learning models. The GitHub repository has more than 5900 stars.

Website Link: https://cvat.org/

Github Link: https://github.com/openvinotoolkit/cvat

ImageTagger

If you are working in a team then this tool is for you. The tool provides an online platform in which you can create a team and label the dataset with the help of different features like point labeling, line, Polygon, bounding boxes, etc. You can also apply the existing annotation to the images. This tool is not so popular but the collaborative image labeling features make it unique. 

Github Link: https://github.com/bit-bots/imagetagger 

ImgLab

Imglab is a simple and fast web-based image annotation tool. The graphical user interface of this tool is very user-friendly and demands very less computational resources. This open-source free image annotation tool provides the ability to draw the rectangle, polygon circles, and feature points on the dataset. The imglab tool can support multiple formats of images like dlib, Pascal VOC, COCO, and Tensorflow. You can use a free-hand draw annotation shape and drag or resize these shapes. The GitHub repository of imglab has 750 plus stars. 

Website Link: https://imglab.in/

Github Link: https://github.com/NaturalIntelligence/imglab

---

Check out more blog posts:

How to Install Python/Anaconda/Jupyter Notebook/Spyder on Ubuntu (Linux)