VID imagenet for object detection

Is there any example of using VID imagenet dataset in pytorch for object detection? especially how to load video frames and annotations.