Labelo’s labeling interface is a powerful and user-friendly environment designed to streamline the process of various types of data annotation, including images, text, and more. This interface provides annotators with a comprehensive set of tools and features to accurately and efficiently create annotations. Key components of the labeling interface include intuitive annotation tools, customizable labeling options, a commenting system, and an annotation history panel.

Labeling Workflow

Labelo provides a structured and intuitive workflow for data annotation, ensuring that users can efficiently and accurately create annotations. It is an efficient data preparation tool for AI. It supports multi-format data annotation.

1. Begin by navigating to the specific task in the project in Labelo where you want to perform annotations.

To begin labeling all tasks in a project, click on the “Label All Tasks” button. If you prefer to label tasks based on their current filter and sort order in the data manager, select the “Label Tasks As Displayed” option instead. This allows you to focus on specific subsets of tasks, enhancing efficiency and organization in your labeling workflow.

2. Choose a label from the pre-created labels.

Once you have navigated to the specific task you wish to annotate, the next step is to select an appropriate label from the list of predefined labels that have been set up for your project. A complete guide on setting up the label can be seen in Configuring Labeling Interface in Labelo.

3. Start labeling by clicking on the image according to the template chosen. 

Apply the label by clicking on the relevant part of the text, image, audio, or other data objects. The system automatically saves your changes. 

4. After completing the labeling, you can submit your annotation by clicking the submit button.

Once you have finished annotating, click the Submit button to finalize the annotation and proceed to the next task. This process ensures that each segment of your data is accurately labeled and systematically recorded.

Now if you want to update the annotation that was already submitted, you can simply open the corresponding task and change the label. Once you made changes, you can click ‘Update’ in order to apply your changes.

Adding Comments

Adding comments to annotations in Labelo is a powerful feature that enhances collaboration, clarity, and quality control in the annotation process. Comments allow annotators and reviewers to communicate effectively, provide context, and address any ambiguities or issues directly within the labeling interface.

In the above-given text area, annotators can add the appropriate comments according to the annotation, and by clicking the icon next to it, it will be sent to all members who can access the particular annotation.

You also have the option to edit, delete, and resolve the comments added.

While annotating data, you can access the commenting feature directly from the annotation view. Labelo provides a separate comments panel where you can view and manage all comments related to a particular data item or annotation.

Annotation History

The annotation history feature in Labelo is an essential tool for tracking and managing the changes made to annotations throughout the labeling process. This feature offers several advantages that enhance the accuracy, accountability, and overall quality of the annotated data. 

In the Labelo interface, there is typically an option to view the annotation history for a specific task or project. This will be accessible via a dedicated button or tab.

Here, when a user tries to update any pre-labeled task, it will be shown as a draft in the annotation history.

Users can browse through the list of changes, each entry showing the details of the edit, including the user who made the change, the time of the change, and any comments.

Label Data as Regions

Labelo offers a powerful feature known as “Regions,” which allows users to label and annotate specific areas within their data. This feature is especially useful for tasks that require precise and detailed annotations, such as image segmentation, object detection, or any scenario where distinct parts of the data need to be highlighted and categorized.

It also includes the feature to lock and hide the labeled regions. Once a labeled region is locked, it cannot be edited further unless it is unlocked. Also, we can hide the annotation so that the labeled region will be removed from the data.

How to Label Overlapping Regions?

When performing image segmentation tasks like labeling with bounding boxes or highlighting text for NLP and NER labeling, you may need to annotate overlapping regions. To simplify this process, you can hide labeled regions after annotating them.

• Select the Label: Choose the label you want to apply to the region.
• Annotate the Region: Draw the bounding box or highlight the text that you want to label.
• Hide the Labeled Region: In the Regions or Labels sidebar, find the region you just labeled and click the eye icon to hide it.
• Label the Overlapping Region: Select the next label for the overlapping region and repeat the process. Draw the bounding box or highlight the text as needed.
• Continue Annotating: Keep hiding and labeling regions until you’ve fully annotated the task. To review your work, click the eye icon next to Regions to hide or show all labeled regions and verify the final result.
• Submit Your Work: Once you’ve completed the annotations, click Submit to finalize and proceed to the next task.

Select Multiple Regions

You can select multiple regions at once when labeling, allowing for simultaneous modifications. To do this, create the desired regions and then hold down the Ctrl key while clicking on each region you wish to include. You can select these regions either directly on the object or within the Regions sidebar. To select a consecutive range of regions in the sidebar, click the first region, then hold Shift and click the last region in the range. Once the regions are selected, you can label or delete them all at once. To deselect a region, simply click on it again or press the “u” key.

How to Duplicate a Region while Labeling?

You can easily duplicate a region to create multiple identical polygons, rectangles, or ellipses. Here’s how to do it:

Select the Region: Begin by selecting the region you want to duplicate.

Duplicate the Region:

Press Ctrl + D to duplicate the region quickly. Alternatively, use the sequence Ctrl + C to copy and Ctrl + V to paste the region.

Placement of Duplicated Regions: The duplicated regions will appear in the same location as the original region.

Adjust the New Region:

Use the arrow tool to move the newly created region to a different location.

Resize the duplicated region as needed to fit your requirements.

By following these steps, you can efficiently create and position multiple identical regions for your labeling tasks.

Change the Label

You can easily modify the label of an existing region in your annotation. Here’s a step-by-step guide on how to do this:

Select the Labeled Region:

Identify the region you wish to re-label. This could be any type of region such as a span, bounding box, image segment, audio region, or other specific area. You can select the region directly on the object or use the Regions sidebar to find and select it.

Choose a New Label:

Once the region is selected, choose the new label from the available options. The system will automatically save your changes as soon as you select the new label.

Submit the Annotation:

After updating the label, click the “Submit” button to finalize your annotation. This will save the changes and allow you to proceed to the next task.

By following these steps, you can efficiently update the labels of your regions, ensuring your annotations remain accurate and up-to-date.

Delete an Annotation

After completing the labeling of a task, you have the option to delete the annotation if necessary. Here’s how you can manage your annotations:

Delete an Annotation:

If you want to remove an annotation from the task, simply click the trash can icon. This will delete the annotation from the task permanently.

Reset Unsaved Annotations:

If you haven’t saved your annotation yet and wish to start over, click the X icon. This action will reset the task, removing all your current labeling activity and allowing you to begin the labeling process anew.

By using these tools, you can efficiently manage your annotations, ensuring that only the desired labels are saved and submitted.

How to delete a Labeled Region?

If you need to delete labeled regions, such as bounding boxes or text spans, follow these steps:

Select the Labeled Region:

Locate the region you want to delete. You can do this either directly on the object or by using the Regions sidebar. Click on the labeled region to select it.

Delete the Region:

Press the Backspace key to remove the selected region immediately. Alternatively, you can navigate to the Results panel. In this panel, find the annotation you wish to delete and remove it from there.

By following these steps, you can easily manage and delete any unwanted labeled regions, ensuring your annotations are accurate and up-to-date.

Skipping a Task

When annotators choose to skip a task, that particular task is removed from their personal labeling queue, meaning they won’t see it again in their list of tasks. However, this skipped task is not removed from the overall labeling queue and remains available for other annotators to work on. This allows the task to still be completed by someone else, ensuring that it eventually receives the necessary attention and annotations. This system helps maintain efficiency and ensures that all tasks are addressed, even if some annotators choose to skip certain tasks.

How to exit from a Labeling flow?

Annotators have the option to exit their labeling flow if they need to pause their annotation work and stop the lead time calculation. Here’s how they can do it:

Submit and Exit:

If annotators want to submit their current work and exit, they can click the drop-down menu next to the Submit button. From the drop-down menu, select either “Submit and Exit” or “Update and Exit.” This action will save and submit their current annotation and then redirect them to the Projects page.

Save Draft and Exit:

If annotators are not finished with their annotation and prefer to save their progress without submitting it, they should first ensure their draft is saved. They can do this by checking the history panel to confirm that the draft has been saved. Once the draft is saved, annotators can navigate directly to the Projects page. This will effectively stop the lead time calculation from running without submitting incomplete work.

These options provide flexibility for annotators to manage their time and workflow efficiently, allowing them to pause and resume their tasks as needed without losing their progress.

Use Keyboard Shortcuts

To enhance efficiency, Labelo provides a variety of keyboard shortcuts that allow users to quickly perform common tasks without relying heavily on mouse navigation. These shortcuts accelerate the annotation process, minimize repetitive actions, and improve overall workflow.

This can be referred to from  Labelo Keyboard Shortcuts in detail.

Features of Outliner

Labelo’s labeling interface includes a feature called Outliner, which accommodates nearly all other labeling features. The Outliner is the section where you can view all the details about annotations, regions, and labeling history.

The outliner panel is collapsible and detachable, allowing users to arrange them according to their wishes. 

In the below-given figure, you can see the button for collapsing the features in the outliner.

Below figure shows the outliner after collapsing.

After collapsing, we can see the option for expanding the outliner from the above figure. By clicking that, we can make the outliner into the previous state.

Similarly the outliner is also detachable which enables the users to place them at any place in the screen. You can simply drag it and drop where you want. Also, you can resize them if you want.

In conclusion, Labelo’s labeling interface offers a robust and intuitive platform for data labeling, equipped with a wide range of tools and features that cater to various annotation needs. By leveraging its customizable options, annotation tools, and support features like commenting and history tracking, annotators can achieve high precision and efficiency in their work, making it an invaluable asset for any data labeling project.