Feature extraction and alignment for navigation applications

An apparatus, comprising:

• a memory storing processor-readable code;
• and
• at least one processor coupled to the memory, the at least one processor configured to execute the processor-readable code to cause the at least one processor to perform operations including:
• determining, based on received sensor data, a first set of features for an area surrounding a vehicle;
• determining, based on an occupancy map for the area surrounding the vehicle, a second set of features for the area surrounding the vehicle, wherein a first subset of the second set of features are each associated with an indication of navigational importance;
• determining, using a query, key, value process, a first attention map for the first set of features and a second attention map for the second set of features, wherein the first attention map is determined based on a first query matrix and a first key matrix, and the second attention map is determined based on a second query matrix and a second key matrix;
• and
• determining, based on the first attention map and the second attention map, a third set of features that align the first set of features with the second set of features, wherein the third set of features align each of at least a second subset of the second set of features with at least one corresponding feature from the first set of features according to a ranking of the first set of features, wherein the ranking is based on whether a feature of the first set of features corresponds to a feature of the at least a second subset that is associated with the indication of navigational importance.

A vehicle, comprising:

• at least one sensor configured to capture sensor data;
• a memory storing processor-readable code;
• and
• at least one processor coupled to the memory and the at least one sensor, the at least one processor configured to execute the processor-readable code to cause the at least one processor to perform operations including:
• determining, based on sensor data from the at least one sensor, a first set of features for an area surrounding the vehicle;
• determining, based on an occupancy map for the area surrounding the vehicle, a second set of features for the area surrounding the vehicle, wherein a first subset of the second set of features are each associated with an indication of navigational importance;
• determining, using a query, key, value process, a first attention map for the first set of features and a second attention map for the second set of features, wherein the first attention map is determined based on a first query matrix and a first key matrix, and the second attention map is determined based on a second query matrix and a second key matrix;
• and
• determining, based on the first attention map and the second attention map, a third set of features that align the first set of features with the second set of features, wherein the third set of features align each of at least a second subset of the second set of features with at least one corresponding feature from the first set of features according to a ranking of the first set of features, wherein the ranking is based on whether a feature of the first set of features corresponds to a feature of the at least a second subset that is associated with the indication of navigational importance.

A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to perform operations comprising:

• determining, based on received sensor data, a first set of features for an area surrounding a vehicle;
• determining, based on an occupancy map for the area surrounding the vehicle, a second set of features for the area surrounding the vehicle, wherein a first subset of the second set of features are each associated with an indication of navigational importance;
• determining, using a query, key, value process, a first attention map for the first set of features and a second attention map for the second set of features, wherein the first attention map is determined based on a first query matrix and a first key matrix, and the second attention map is determined based on a second query matrix and a second key matrix;
• and
• determining, based on the first attention map and the second attention map, a third set of features that align the first set of features with the second set of features, wherein the third set of features align each of at least a second subset of the second set of features with at least one corresponding feature from the first set of features according to a ranking of the first set of features, wherein the ranking is based on whether a feature of the first set of features corresponds to a feature of the at least a second subset that is associated with the indication of navigational importance.