| aeroplane": 1, "bicycle": 2, "bird": 3, "boat": 4, "bottle": 5, "bus": 6, "car": 7, |
|---|
What is Auto Segmentation.
Try Auto Segmentation at Labellerr .
Auto segmentation is a process of automatically identifying and delineating objects in an image. It is achieved via using a machine learning algorithm, which is trained on manually segmented data to accurately identify objects in the image. Auto segmentation can be used to speed up the process of data annotation. By using a machine learning algorithm to accurately identify objects in an image, the need for manual annotation is reduced.
We have used the capability of AUTO SEGMENTATION to build a more accurate, fast & sophisticated annotation pipeline. This feature enables a user to draw segmentation around a object via only drawing BOUNDING BOXES .
NOTE:- Before going forward, please ensure that your are familiar with the Annotation Projects, Datasets and Annotation Pipelines at Labellerr. If not please visit the below links :-
A. First we need to create a polygon type annotation question with the name of object that we want to auto-segment. Select a segmentation object. Selecting the object an auto segment icon in the tools section will appear. Clicking on the auto segment icon , the user will be able to draw a bounding box over the required object in the image. After the drawing the Bounding Box, the object will be segmented.
Note:- The solution supports the only following 20 classes. So the polygon type questions must have the same class-name as their labels.