Abstract: Training object class detectors typically requires a large set of images with objects annotated by bounding boxes. However, manually drawing bounding boxes is very time consuming. In this ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results