Bulk Actions

Select multiple assets in your dataset and delete them all at once using the bulk actions toolbar in Datature Vi.

Bulk actions in Datature Vi let you select multiple assets and delete them in a single step. This is the fastest way to clear out test data, remove duplicates, or clean up an entire category of poor-quality images.

Before You Start
  • You need an existing dataset with assets you want to remove.
  • You need edit access to the dataset (Owner or Editor role).
  • Download a backup if there is any chance you will need the deleted assets later.

Once confirmed, all selected assets and every annotation on those assets are immediately and permanently deleted. There is no undo and no recycle bin. The only way to restore deleted assets is to re-upload from a local backup.

1

Open the Dataset Explorer

Open the Dataset Explorer

Click the Explorer tab in the sidebar to list the assets in your dataset. .

You should see
Dataset Explorer after bulk deletion showing no assets

The dataset is now empty if all assets were selected in the delete operation.

After the deletion

Once the bulk delete completes:

  • All selected assets are removed from the Dataset Explorer.
  • The total asset count decreases by the number of deleted assets.
  • The selection toolbar disappears.
  • The dataset view refreshes to show the remaining assets.

Check the dataset overview to confirm the updated asset count and annotation distribution.

Troubleshooting

  • Click the checkbox area itself, not the asset thumbnail image.
  • Refresh the page if checkboxes remain unresponsive.
  • Confirm you are on the Assets tab (not Annotations or Insights).
  • The toolbar only appears after you select at least one asset.
  • Make sure your browser window is wide enough to display the full toolbar.
  • Bulk deletion cannot be undone through the interface.
  • Re-upload from your local backup if you downloaded the dataset before deleting.
  • Re-annotate if you have the original image files but not the annotations.

Do this with the Vi SDK

import vi

client = vi.Client(
    secret_key="your-secret-key",
    organization_id="your-organization-id"
)

response = client.datasets.bulk_delete_assets(
    dataset_id="your-dataset-id",
    filter_criteria='{"status": "error"}'
)

For more details, see the full SDK reference.

Next steps

Delete An Asset

Remove individual assets one at a time using the action menu.

Dataset Overview

Check the updated asset count and annotation distribution after cleanup.

Download Full Dataset

Create a backup before your next round of deletions.