Detection and classification of traffic signs using camera-radar fusion

A method comprising:

• generating, by a data processing system of a vehicle, for each of a plurality of times:
• a respective camera feature of a plurality of camera features, the respective camera feature characterizing one or more camera images of an environment associated with a respective time of the plurality of times, and
• a respective radar feature of a plurality of radar features, the respective radar feature characterizing one or more radar images of the environment associated with the respective time of the plurality of times;
• fusing, by the data processing system, the plurality of camera features with the plurality of radar features across the plurality of times to obtain a fused tensor;
• processing, by the data processing system using a neural network (NN), the fused tensor to obtain semantic content of one or more traffic signs in the environment;
• and
• controlling, by a driving control system of the vehicle, the vehicle based on the semantic content of the one or more traffic signs.

A non-transitory computer-readable medium storing instructions that, when executed by a processing device, cause the processing device to:

• generate, for each of a plurality of times:
• a respective camera feature of a plurality of camera features, the respective camera feature characterizing one or more camera images of an environment associated with a respective time of the plurality of times, and
• a respective radar feature of a plurality of radar features, the respective radar feature characterizing one or more radar images of the environment associated with the respective time of the plurality of times;
• fuse the plurality of camera features with the plurality of radar features across the plurality of times to obtain a fused tensor;
• process, using a neural network (NN), the fused tensor to obtain semantic content of one or more traffic signs in the environment;
• and
• cause a driving control system of a vehicle to control the vehicle based on the semantic content of the one or more traffic signs.