Approaches for encoding environmental information
Systems, methods, and non-transitory computer-readable media can access a plurality of schema-based encodings providing a structured representation of an environment captured by one or more sensors associated with a plurality of vehicles traveling through the environment. The plurality of schema-based encodings can be clustered into one or more clusters of schema-based encodings. At least one scenario associated with the environment can be determined based at least in part on the one or more clusters of schema-based encodings.
1. A computer-implemented method comprising:
encoding, by a computing system, metadata for a map region based on map features that describe agents, actions performed by the agents at different times over a period of time, and environment associated with the map region;
determining, by the computing system, one or more scenarios for the map region based on the metadata;
determining, by the computing system, one or more first frequencies for the one or more scenarios in the map region;
identifying, by the computing system, a second map region associated with one or more second frequencies for the one or more scenarios, wherein the one or more second frequencies are within one or more threshold frequency differences of the one or more first frequencies; and
generating, by the computing system, a risk profile for the map region based on the one or more scenarios, the metadata, and a risk profile for the second map region, wherein the risk profile includes a first level of risk of encountering people and a second level of risk of encountering non-moving objects.
2. The computer-implemented method of claim 1 , wherein the map features include zone information of a road segment in the map region, a road segment quality of the road segment, and contextual information of the road segment.
3. The computer-implemented method of claim 1 , further comprising:
modifying, by the computing system, operation of a vehicle in the map region based on the risk profile for the map region.
4. The computer-implemented method of claim 1 , wherein the determining the one or more scenarios comprises:
generating, by the computing system, one or more histograms that represent one or more frequencies associated with one or more scenario families that include the one or more scenarios.
5. The computer-implemented method of claim 1 , wherein the one or more scenarios are associated with one or more levels of difficulty for navigation, and wherein the risk profile for the map region is based on the one or more levels of difficulty for navigation.
6. The computer-implemented method of claim 1 , wherein the risk profile for the map region is based on risk profiles of other map regions that satisfy a threshold similarity with respect to frequencies of scenario families for the map region.
7. The computer-implemented method of claim 1 , wherein the one or more scenarios are associated with a set of predefined scenario families that correspond with interactions of objects detected at the map region.
8. The computer-implemented method of claim 1 , wherein the metadata includes a time-based representation of objects detected over the period of time at the map region and information obtained from a semantic map associated with the map region.
9. The computer-implemented method of claim 1 , wherein the metadata is encoded based on a predefined scenario schema, and wherein the predefined scenario schema provides a structured representation of the map features associated with the map region.
10. A system comprising:
at least one processor; and
a memory storing instructions that, when executed by the at least one processor, cause the system to perform operations comprising:
encoding metadata for a map region based on map features that describe agents, actions performed by the agents over a period of time, and environment associated with the map region;
determining one or more scenarios for the map region based on the metadata;
determining one or more first frequencies for the one or more scenarios in the map region;
identifying a second map region associated with one or more second frequencies for the one or more scenarios, wherein the one or more second frequencies are within one or more threshold frequency differences of the one or more first frequencies; and
generating a risk profile for the map region based on the one or more scenarios, the metadata, and a risk profile for the second map region, wherein the risk profile includes a first level of risk of encountering people and a second level of risk of encountering non-moving objects.
11. The system of claim 10 , wherein the map features include zone information of a road segment in the map region, a road segment quality of the road segment, and contextual information of the road segment.
12. The system of claim 10 , the operations further comprising:
modifying operation of a vehicle in the map region based on the risk profile for the map region.
13. The system of claim 10 , wherein the determining the one or more scenarios comprises:
generating one or more histograms that represent one or more frequencies associated with the one or more scenarios.
14. The system of claim 10 , wherein the one or more scenarios are associated with one or more levels of difficulty for navigation, and wherein the risk profile for the map region is based on the one or more levels of difficulty for navigation.
15. A non-transitory computer-readable storage medium including instructions that, when executed by at least on processor of a computing system, cause the computing system to perform operations comprising:
encoding metadata for a map region based on map features that describe agents, actions performed by the agents over a period of time, and environment associated with the map region;
determining one or more scenarios for the map region based on the metadata;
determining one or more first frequencies for the one or more scenarios in the map region;
identifying a second map region associated with one or more second frequencies for the one or more scenarios, wherein the one or more second frequencies are within one or more threshold frequency differences of the one or more first frequencies; and
generating a risk profile for the map region based on the one or more scenarios, the metadata, and a risk profile for the second map region, wherein the risk profile includes a first level of risk of encountering people and a second level of risk of encountering non-moving objects.
16. The non-transitory computer-readable storage medium of claim 15 , wherein the map features include zone information of a road segment in the map region, a road segment quality of the road segment, and contextual information of the road segment.
17. The non-transitory computer-readable storage medium of claim 15 , the operations further comprising:
modifying operation of a vehicle in the map region based on the risk profile for the map region.
18. The non-transitory computer-readable storage medium of claim 15 , wherein the determining the one or more scenarios comprises:
generating one or more histograms that represent one or more frequencies associated with the one or more scenarios.