Kili Technology allows to leverage all object detection tools (point, line, polygon, bounding box, segmentation) on multi-frames interface for video. It makes it easy to create, edit, modify objects on one frame or across frames. All combinations with classification and nested classification can be leveraged.
You can upload videos using a csv file with two columns : one for the external identifier, and the other containing a link to the video. Videos are converted to an array of images. In order to have a fluid interface, try to work on videos of less than 1000 frames.
We advise you to split long videos into shorter ones with up to 1000 frames, for performance reasons. As performance also depends on the number of annotations, we recommend having a total maximum of 10000 annotations accross all frames for each asset.
Uploading list of images
We allow you to upload videos as list of extracted images with the API (see the recipe). This can be useful for viewing 3D volumes of image data. We show you how to do that here.
This allows you to upload videos with up to 20000 images. If you want more, please contact us.
Object detection tasks
Create an object on one frame
By default, when you draw an object on one frame, it is propagated to all following frames, until the last frame of the video.
Delete an object on one frame
When you delete an object on a single selected frame, it will be deleted only on the selected frame.
Create an object on multiple frames
To create an object on a restrained range of frames, select the start or the end frame, click on Shift, and click on the final frame. The whole section will be selected. Then create the object on the section.
Delete an object on multiple frames
To delete an object on a restrained range of frames, select the start or the end of the range by clicking on a frame, pressing and holding Shift and clicking on another frame. All the frames in between are going to be selected. Then, click on the object you want to destroy on the image and press backspace to delete it.
Edit an object on multiple frames
You can edit and modify an object on one frame. Then:
- the modification will be applied to all following frames;
- the frame will be turned into a "key frame" and marked with a lock icon on the interface;
When modified, bounding box and point annotations are interpolated between the preceding key frame and the current key frame, as well as between the current key frame and the following key frame. If there are no preceding/following key frame, then there is no interpolation.
For bounding box jobs, you can use a tracker to help you. It automatically labels an object, by following it.
- Set it up by ticking the checkbox after creating the bounding box job, or add it to your existing job in the json interface.
- Use it in the application : click on a bounding box category, and press the tracking button before drawing the bounding box.
For Point, choose tool: Point.
For Line, choose tool: Line.
For Polyon, choose tool: Polygon.
For Bounding box, choose tool: Bounding box.
For Segmentation, choose tool: Semantic.
Shortcuts are provided to help you annotate quickly :
- A shortcut will be generated for each different class of object to detect : see shortcuts section
- On the frame interface, you can use PLAY and PAUSE buttons to play and pause the video, or their common shortcut SPACE.
- To scroll quickly in the frames, you can also use SPACE + wheel of the mouse (or 2 fingers on the pad).
- You also have zoom shortcuts available.
- You can also edit faster with this shortcut to correct the last segments drawn for the current object, or to escape the editing interface
- To annotate on a frame without propagating the annotation on the full video, use Ctrl + click to select the frame.