I’m currently labelling a bunch of videos with the Video Object Detection and Tracking template, and I appreciate the additional precision the Zoom tool provides, but the “ZoomPanTool” from the Image template is not available, and when zooming close to the edge of a frame, the Zoom tool does NOT go “out of bounds”, so I cannot center the edge of a frame in the editor, and I find it VERY annoying.
What I have difficulty understanding is, considering a video frame is basically an image, why the Video’s Zoom tool behaves differently from the Image’s Zoom tool. Couldn’t the same tool be used in both templates?
I’d gladly implement this feature in the Video OD&T’s tools. However, I don’t have a lot of experience in development, let alone web development, and this is my first exposure to React, so I find it quite difficult to understand how the code works.
Setup
Using Docker to run the application.
Label Studio Docker Image: 20240711.125956-ls-release-1-13-0-d11e5ebf6