When the application draws prediction results(labels, bounding boxes, descriptions) on the output video ,the overlays can become large enough to cover important parts of the scene. It would be useful to have an option to make these semi-transparent, so the underlying video remains visible for observability while still showing the prediction information.
When the application draws prediction results(labels, bounding boxes, descriptions) on the output video ,the overlays can become large enough to cover important parts of the scene. It would be useful to have an option to make these semi-transparent, so the underlying video remains visible for observability while still showing the prediction information.