Identifying objects for display in a situational-awareness view of an autonomous-vehicle environment

1. A method comprising, by one or more computing devices:

receiving sensor data captured by a sensor array of an autonomous vehicle, wherein the sensor data indicates an object in an external environment of the autonomous vehicle;

accessing secondary data associated with the object from one or more secondary-data sources that are independent of the sensor array of the autonomous vehicle;

determining, based on at least the sensor data and the secondary data, a classification of the object and an associated confidence score for the classification;

determining, based on at least the classification of the object and the associated confidence score, an object graphic to visually represent the object; and

causing a visual representation of the external environment of the autonomous vehicle including the determined object graphic to be presented at a display associated with the autonomous vehicle.

2. The method of claim 1 , wherein determining the object graphic comprises:

identifying, from the sensor data, one or more data points that correspond to the object;

determining, from the secondary data, one or more features of the object; and

determining, by applying a machine-learning model to the one or more data points and the one or more features, a confidence score that indicates a probability that the one or more data points correspond to a particular object type, wherein

the object graphic determined to visually represent the object is further based on at least the confidence score that indicates the probability that the one or more data points correspond to the particular object type.

3. The method of claim 1 , further comprising, before accessing the secondary data associated with the object, determining that the sensor data only partially describes the object.

4. The method of claim 3 , wherein determining that the sensor data only partially describes the object comprises determining that a similarity score between one or more data points corresponding to the object and a pre-determined pattern corresponding to a pre-classified object is below a pre-determined threshold.

5. The method of claim 1 , wherein determining the object graphic to visually represent the object further comprises:

determining, based on at least the sensor data, one or more features of the object, the one or more features comprising a color, a shape, a size, or a gait associated with the object; and

causing the object graphic to be presented with one or more of the one or more features.

6. The method of claim 1 , wherein the secondary data is user-generated data submitted by one or more users of a dynamic transportation matching system.

7. The method of claim 1 , further comprising:

determining a confidence score for the classification of the object that represents a probability that the classification is correct;

determining that the confidence score for the classification of the object is below a predetermined threshold level; and

causing a request for a passenger of the autonomous vehicle to input a user-provided classification for the object to be presented at the display associated with the autonomous vehicle.

8. The method of claim 1 , further comprising:

determining that a confidence score associated with a classification of the object based on the sensor data captured by the sensor array of the autonomous vehicle alone is below a threshold score; and

accessing the secondary data to a update the classification of the object.

9. The method of claim 1 , wherein the secondary data is map data stored by the one or more computing devices, and wherein the map data comprises a street name, a latitude and longitude coordinate, a region, or an intersection associated with the object.

10. The method of claim 9 , further comprising inputting the map data into a machine-learning model that uses the map data to determine a probability that the object has a particular classification.

11. The method of claim 1 , wherein the visual representation is presented on a computing device of the autonomous vehicle.

12. The method of claim 1 , wherein the visual representation is presented on a client device of a current passenger of the autonomous vehicle.

13. The method of claim 1 , wherein the secondary data is historical data comprising a plurality of past classifications made by the one or more computing devices during one or more previous trips of the autonomous vehicle, wherein the plurality of past classifications is used by a machine-learning model to determine a probability that the object has a particular classification.

14. The method of claim 1 , wherein the display associated with the autonomous vehicle provides a situational-awareness view for a passenger of the autonomous vehicle.

15. A system comprising: one or more processors and one or more computer-readable non-transitory storage media in communication with the one or more processors, the one or more computer-readable non-transitory storage media comprising instructions that, when executed by the one or more processors, are configured to cause the system to perform operations comprising:

receiving sensor data captured by a sensor array of an autonomous vehicle, wherein the sensor data indicates an object in an external environment of the autonomous vehicle;

accessing secondary data associated with the object from one or more secondary-data sources that are independent of the sensor array of the autonomous vehicle;

determining, based on at least the sensor data and the secondary data, a classification of the object and an associated confidence score for the classification;

determining, based on at least the classification of the object and the associated confidence score, an object graphic to visually represent the object; and

causing a visual representation of the external environment of the autonomous vehicle including the determined object graphic to be presented at a display associated with the autonomous vehicle.

16. The system of claim 15 , wherein determining the object graphic comprises:

identifying, from the sensor data, one or more data points that correspond to the object;

determining, from the secondary data, one or more features of the object; and

determining, by applying a machine-learning model to the one or more data points and the one or more features, a confidence score that indicates a probability that the one or more data points correspond to a particular object type, wherein

the object graphic determined to visually represent the object is further based on at least the confidence score that indicates the probability that the one or more data points correspond to the particular object type.

17. The system of claim 15 , wherein the instructions are further configured to cause the system to perform further operations comprising, before accessing the secondary data associated with the object, determining that the sensor data only partially describes the object.

18. One or more computer-readable non-transitory storage media including instructions that, when executed by one or more processor of a computing system, are configured to cause the one or more processors to perform operations comprising:

receiving sensor data captured by a sensor array of an autonomous vehicle, wherein the sensor data indicates an object in an external environment of the autonomous vehicle;

accessing secondary data associated with the object from one or more secondary-data sources that are independent of the sensor array of the autonomous vehicle;

determining, based on at least the sensor data and the secondary data, a classification of the object and an associated confidence score for the classification;

determining, based on at least the classification of the object and the associated confidence score, an object graphic to visually represent the object; and

causing a visual representation of the external environment of the autonomous vehicle including the determined object graphic to be presented at a display associated with the autonomous vehicle.

19. The one or more computer-readable non-transitory storage claim 18 , wherein determining the object graphic comprises:

identifying, from the sensor data, one or more data points that correspond to the object;

determining, from the secondary data, one or more features of the object; and

determining, by applying a machine-learning model to the one or more data points and the one or more features, a confidence score that indicates a probability that the one or more data points correspond to a particular object type, wherein

the object graphic determined to visually represent the object is further based on at least the confidence score that indicates the probability that the one or more data points correspond to the particular object type.

20. The one or more computer-readable non-transitory storage claim 18 , wherein the instructions are further configured when executed to cause the one or more processors to perform further operations comprising, before accessing the secondary data associated with the object, determining that the sensor data only partially describes the object.