Skip to content


Images panel shows all images of current dataset.



  1. Total number of images in dataset
  2. Filter of images
  3. Preview picture
  4. File name
  5. Number of annotations
  6. Last modification date (i.e. modification of annotation)
  7. Attach new tag to an image
  8. Download original image file
  9. Delete image and all of its annotations

Change selected image

Current image is highlighted in blue color. You can switch between images by clicking on file names in Images Panel or using Navigation Panel or using Left Arrow and Right Arrow hotkeys.


Images Filter (50%)

If there are a lot of images in dataset, you can use filter popup to leave only images that pass certain conditions or order images.

Open filter popup by clicking Filter icon (see overview above)


Create tags like Valid and Invalid to mark images that have correct / incorrect annotations and then apply filter to quickly find images that need additional labeling.


Choose Objects count - None to quickly find images that need annotations.


Image Tags (70%)

Tags are predefined list of keywords to quickly mark images. Tags can be used in images filter, DTL or downloaded along with JSON annotation files.

You can add new image tags on Project page under Tags/Image tags tab in dashboard or right in annotation module using popup displayed on this image above.


Assign a hotkey when you add a new tag. Thus you can quickly attach tag to an image by pressing corresponding hotkey on a keyboard.


Create tag Valid and assign it hotkey V. Now you can mark images with correct annotation by pressing V key on a keyboard and go to the next image by pressing Right Arrow key.



Number of images shown in panel is limited to 50. If current dataset contains more images than that, pagination will be shown below.

Image size & resolution

Don't worry about image size. After uploading dataset using Import, apart from original images, we generate two more versions: 40x40 compressed preview for Images Panel and compressed version of original that is used for showing in annotation module. Original is used in DTL and in downloads.