Implement/Get an model for object detection (even better if it can do instance seg like MaskRCNN)
We need it to include at least classes ['vehicle', 'pedestrian', 'sign', 'bicycle'], everything else are just bonus.
Note that great transfer learning stuff is up for grab here .
Implement/Get an model for object detection (even better if it can do instance seg like MaskRCNN)
We need it to include at least classes
['vehicle', 'pedestrian', 'sign', 'bicycle']
, everything else are just bonus. Note that great transfer learning stuff is up for grab here .