The predictions from monoflex was passed on to AB3DMOT, and the bounding boxes and velocity was visualized as shown in figure 12. In most of the cases the 3D bounding box prediction is successful and the corresponding velocity was visualized on top of the bounding box. There were few cases when the 3D bounding box visualization failed and the corresponding velocity was unable to be predicted. The monoflex baseline was also tested on the TuSimple dataset and the 2D bounding box was obtained from the predicted 3D bounding box by considering the minimum and maximum x and y coordinate in the 3D keypoints. The figure 13 shows successful detections and failure detections. As shown in the figure the detection fails to detect cars which are far away from the ego centric vehicle but for cars which under the 15m inter- vehicle distance the detections are successful.