Consensus with SAM 2 - Images and Videos

Annotation of images and videos depends on your Ontology. Our Ontology E2E Fruit Ontology has the following label types:

• Bounding boxes
• Polygons
• Bitmasks
• Classifications

In this section, the following Collaborators roles are shown:

• Annotators labeling data
• Reviewers reviewing labels created by Annotators
• Team Manager managing the Annotators and Reviewers
• Project Admin managing the Project and exporting labels

1. Prepare to Label

   Team Manager or Project Admin

   The Team Manager or Project Admin can prioritize certain data to be labeled and reviewed first. Let’s prioritize the blueberries, cherries, and apples to be labeled first by setting the priority for those files to 75.

   Access Project

   

   Set Priority to 75

   

   Annotators

   Annotators can configure the Label Editor so they can more effectively and efficiently label data.

   

2. Use SAM 2 to Label Data

   Team Manager or Project Admin

   The Team Manager or Project Admin can monitor the performance and progress of the annotation team.

   

   Annotators

   This is a Consensus Project using the default Consensus workflow Workflow with the default values (3 Annotators). This means that 3 Annotators are REQUIRED to annotate and submit a data unit before the data unit moves to the REVIEW stage.

   Annotators use SAM 2 or manually annotate fruit in the Dataset using object labels. Classifications are manually annotated. We’ll focus on blueberries, cherries, apples and whether the fruit is ripe.

   Use hotkeys to speed up your annotation process.

   Bounding Boxes

   We’ll use SAM 2 and manually created bounding boxes to label cherries.

   Images

   • Annotator 1 - Using SAM 2

   

   • Annotator 2 - Manual annotation

   

   • Annotator 3 - Using SAM 2

   

   Videos

   • Annotator 1 - Using SAM 2

   

   • Annotator 2 - Manual annotation

   

   • Annotator 3 - Using SAM 2

   

   Polygons

   We’ll use SAM 2 and manually created polygons to label blueberries.

   Images

   • Annotator 1 - Using SAM 2

   

   • Annotator 2 - Manual annotation

   

   • Annotator 3 - Using SAM 2

   

   Videos

   • Annotator 1 - Using SAM 2

   

   • Annotator 2 - Manual annotation

   

   • Annotator 3 - Using SAM 2

   

   Bitmasks

   We’ll use SAM 2 and manually created bitmasks to label apples.

   Images

   • Annotator 1 - Using SAM 2

   

   • Annotator 2 - Manual annotation

   

   • Annotator 3 - Using SAM 2

   

   Videos

   • Annotator 1 - Using SAM 2

   

   • Annotator 2 - Manual annotation

   

   • Annotator 3 - Using SAM 2

   

   Classifications

   We’ll label using the Ripe? classification.

   Images

   • Annotator 1

   

   • Annotator 2

   

   • Annotator 3

   

   Videos

   • Annotator 1

   

   • Annotator 2

   

   • Annotator 3

   

3. Review Labels

   Team Manager or Project Admin

   The Team Manager or Project Admin can monitor the performance and progress of the review team.

   

   Review Labels

   Reviewers review object labels and classifications from multiple Annotators and decide on the best object labels and classifications. We’ll focus on blueberries, cherries, apples (object labels) and whether the fruit is ripe (classifications).

   You can approve labels/classifications on a task one at a time (from the left panel) or all at once (using the Approval all button).

   When there is an issue with labels/classifications, Reviewers can:

   • Reject the task and add a comment about why a task was rejected. Rejected tasks go back to the person who added the labels/classifications.
   • Edit labels directly using the Edit labels button and then approve the task.

   Rejected Tasks

   Before going over tasks that you can approve, we’ll look at one that will be rejected. The reason this task will be rejected is that the image contains juniper berries and not blueberries.

   

   Bounding Boxes

   Images

   

   Videos

   

   Polygons

   Images

   

   Videos

   

   Bitmasks

   Images

   

   Videos

   

   Classifications

   Images

   

   Videos

   

4. Export Labels

   Only Project Admins can export labels from Encord.

   Project Admin

   Export the labels from the Project Labels page.