For the sake of simplicity, for the time being, we have just been using a visualization of the ground truth annotations from the nuScenes dataset.
However, in the future, we ideally want live-predictions from the model itself. Therefore, see if we can get live predictions from either the patched model repository package or the TensorRT ROS wrapper version
For the sake of simplicity, for the time being, we have just been using a visualization of the ground truth annotations from the nuScenes dataset.
However, in the future, we ideally want live-predictions from the model itself. Therefore, see if we can get live predictions from either the patched model repository package or the TensorRT ROS wrapper version