Figure 4
A diagram showing a three D Foreground voxel prediction network block and a S S F T module block.

The visualization of highly occluded objects detections on KITTI val set. Note that the green, yellow and cyan boxes indicates the class of car, cyclist and pesdestrian, respectively. Also, the red circle denotes the main differences amoing the models’ detection.

or Create an Account

Close subscription notice
Close access options