FAQ

After you click on the invitation link received via email, you’ll be redirected to SuperAnnotate, where you can see the team name in the top left corner of the page. If you have multiple teams, you can select the one you want from the dropdown menu. See here.

Go to Billing and check the users count, then go to Team Settings and check the number of team members. Note that users with pending invitations are also counted as team members.

If you don’t want to manually assign items to Annotators and QAs, ask them to click Request Tasks in Data. They'll be automatically assigned up to the number of items that's set in Request item number in Settings. The minimum number is 1, and the maximum number is 50 for Video Projects and Document Projects and 200 for Image Projects.

Project Admins can’t upload or download data unless you give them permission from Settings.

By default, the projects are sorted by creation date (newest to oldest). To change the sorting direction, click on the ellipsis (...) in the search bar, then click Sort Direction. You can also sort the projects by type and name. To do that, click on the ellipsis (...) in the search bar and click Type or Name.

Yes, you can copy items from one folder to another. To do that, select the item, click Copy, and select the folder where you want to copy the item. You can also do this with SDK.

SuperAnnotate supports a two-level folder structure. Note that if the integrated storage has a folder structure with more than two levels, only the first-level and second-level folders will be uploaded.

You can attach an instructions document or link to a project in Settings or in the Editor. Learn more.

The SDK token will automatically expire after 1 year from the date of generation. However, you can revoke it at any time.

Only Team Owners have access to their team’s token for Python SDK.

SuperAnnotate supports conversions from common annotation formats like COCO, Pascal VOC, and YOLO.
Check the table below for the list of supported formats and task combinations in the import_annotation function.

Format	Project type	Task
COCO	Vector	instance_segmentation
COCO	Vector	object_detection
COCO	Vector	keypoint_detection
COCO	Pixel	panoptic_segmentation
VOC	Vector	instance_segmentation
VOC	Vector	object_detection
YOLO	Vector	object_detection

Check the table below for the list of supported formats and task combinations in the export_annotation function.

Format	Project type	Task
COCO	Pixel	panoptic_segmentation
COCO	Pixel	instance_segmentation
COCO	Vector	instance_segmentation
COCO	Vector	object_detection
COCO	Vector	keypoint_detection

• Image size: 100MB
• Vector image resolutions: 100MP
• Pixel image resolutions: 4MP

• Images (JPG, JPEG, PNG, WEBP, TIFF, BMP, and some types of TIF.)
• Videos (MP4, AVI, MOV, FLV, MPEG, and WEBM.)
• Documents (TXT files with UTF encoding only.)

This table gives you an overview of how to import images, videos, and documents to SuperAnnotate. For more information:

• Images
• Videos
• Documents

Item	Import from computer	Import from S3 bucket	CLI upload	Attach URLs	Integrations
Images	Yes	Yes	Yes	Yes	Yes
Videos	Yes (Vector Projects)	No	Yes (Pixel Projects and Vector Projects)	Yes (Video Projects)	Yes
Documents	No	No	No	Yes	Yes

You can upload annotations from a local storage or a cloud storage using Python SDK. Learn more.

• From the UI: up to 2,000 items
• Via Python SDK: up to 50,000 items

Each item can only have 1 Annotator and 1 QA assigned to it.

No, Annotators and QAs can only see the items that are assigned to them. Note that they can’t request more images than the limit you set in Settings. Learn more about project permissions.

Your data is in safe hands. Here's how we do it:

• We use AWS Cognito services and keep your data secure in S3 buckets. Only your team has access to your data.
• We do not provide any public links. So no one can access your data after you export it.
• If you request to delete your data, we will delete it the same day.
• You can restrict your team members from downloading your data. Learn more about team permissions.
• We sign a mutual NDA when you want to hire annotators through our platform.

There are three main differences between uploading images in the UI and via SDK:

• Bulk uploading through the SDK is faster than uploading in the UI.
• You can upload up to 2,000 items at a time in the UI, while you can upload up to 50,000 items per folder at a time via SDK.
• You can attach URLs from cloud-based storages to your projects via Python SDK.

You can download your annotations in one of the following formats: SuperAnnotate or COCO. You can use converters to convert your annotations from COCO to other formats. Learn more about export formats.

You can’t recover deleted items. Note that when you upload items then delete them, they’ll still be counted towards the number of item uploads.

Yes, you can create an AWS integration to add items from your AWS S3 bucket to your project. SuperAnnotate will have read-only access to your data; your items are stored in your AWS S3 buckets and are only displayed in SuperAnnotate. Learn more.

Only Team Owners and Team Admins can attach items using integrations. You can allow Project Admins to attach items from integrated storages by enabling Upload in Settings.

• You can't upload more than 50,000 items to the root or a folder. The remaining items will be skipped.
• A project can't contain more than 500,000 items.
• A folder name can't exceed 120 characters and can't contain the following characters: / \ : * ? " < > |

It's up to you to decide your S3 bucket's region. Note that having the S3 in the region where your team is working will improve the image loading latency.

The Image (Legacy) Editor allows you to complete semantic and panoptic segmentation using edge detection and image segmentation algorithms. The Image (Legacy) Editor includes tools such as Smart segmentation, Brush, and Polygon. The Image Editor allows you complete object detection projects. It includes tools such as Points, Polyline, Polygon, Bounding Box, Ellipse, Template, Cuboid, and many more.

Select the polygon tool, hold down the left click, and outline your object.

You can add point labels to polylines, polygons, bounding boxes, rotated boxes, ellipses, and cuboids. To add a point label:

1. Select an instance.
2. Right-click on a point.
3. Enter a label name.
4. Select the check icon.

You can press Tab to shift between different instances.

The Neural Network is an environment where you can create and train prediction models. Once you’ve created a model, you can use Smart Prediction to predict annotations. Learn more.

You can run a training only using images with the Completed status. Please check your images' statuses.