Adding tags to sensor data via a plurality of models and querying the sensor data
Provided are methods for customized tags for annotating sensor data, which can include receiving sensor data captured during a plurality of sensor data capture sessions, processing the sensor data using a plurality of machine learning models to identify a plurality of capture session collections represented in the sensor data, filtering the sensor data based at least partly on a user-specified category of the plurality of categories of capture session to identify a capture session collection, of the plurality of capture session collections, representing sensor data of one or more sensor data capture sessions that conforms to the user-specified category, and transmitting the sensor data of one or more sensor data capture sessions that conforms to the user-specified category to an end user computing device. Systems and computer program products are also provided.
1. A method comprising:
receiving, with at least one processor, sensor data captured by at least one sensor;
processing, with the at least one processor, the sensor data using a plurality of machine learning models to identify a plurality of capture session collections represented in the sensor data, a machine learning model of the plurality of machine learning models configured to identify at least a portion of the sensor data as conforming to a category of capture session from a plurality of categories of capture session; and
filtering, with the at least one processor, the sensor data based at least partly on a user-specified category of the plurality of categories of capture session to identify a capture session collection, of the plurality of capture session collections, representing at least a portion of the sensor data classified as conforming to the user-specified category.
2. The method of claim 1 , wherein the sensor data is captured during a plurality of sensor data capture sessions.
3. The method of claim 1 , wherein the sensor data is captured by at least one sensor during at least one vehicle trip.
4. The method of claim 1 , wherein a second machine learning model of the plurality of machine learning models is configured to identify at least a second portion of the sensor data as conforming to a second category of capture session from the plurality of categories of capture session.
5. The method of claim 1 further comprising transmitting, with the at least one processor, the at least a portion of the sensor data classified as conforming to the user-specified category to a computing device.
6. The method of claim 1 , wherein the at least a portion of the sensor data classified as conforming to the user-specified category corresponds to a plurality of sensor data capture sessions.
7. The method of claim 1 , wherein the sensor data comprises at least one of camera data, lidar data, or radar data.
8. The method of claim 1 , wherein the at least one sensor comprises at least one of a camera image sensor, a lidar sensor, or a radar sensor.
9. The method of claim 1 further comprising:
routing, with the at least one processor, the at least a portion of the sensor data classified as conforming to the user-specified category for annotation to generate annotated sensor data, wherein the annotated sensor data comprises an annotation associated with an object within an image; and
training, with the at least one processor, an additional machine learning model based at least in part on the annotated sensor data.
10. The method of claim 1 further comprising:
causing, with the at least one processor, display of the at least a portion of the sensor data classified as conforming to the user-specified category via a display of a computing device;
receiving, with the at least one processor, a confirmation from the computing device based at least in part on causing display of the at least a portion of the sensor data classified as conforming to the user-specified category; and
routing, with the at least one processor, the at least a portion of the sensor data classified as conforming to the user-specified category for annotation to generate annotated sensor data based at least in part on receiving the confirmation from the computing device.
11. The method of claim 1 , wherein the at least one sensor is associated with a vehicle, wherein the user-specified category identifies a particular status of the vehicle, wherein the at least a portion of the sensor data classified as conforming to the user-specified category is associated with the particular status of the vehicle, and wherein the particular status of the vehicle comprises at least one of a speed of the vehicle, an acceleration of the vehicle, an orientation of the vehicle, a location of the vehicle, or a position of the vehicle.
12. The method of claim 1 , wherein the user-specified category identifies a particular environmental characteristic, wherein the at least a portion of the sensor data classified as conforming to the user-specified category is associated with the particular environmental characteristic, and wherein the particular environmental characteristic comprises at least one of lighting, weather, or time of day.
13. The method of claim 1 , wherein the user-specified category identifies a particular type of object, wherein the at least a portion of the sensor data classified as conforming to the user-specified category is associated with the particular type of object, and wherein the particular type of object comprises at least one of a bicycle, a vehicle, or a pedestrian.
14. The method of claim 13 , wherein filtering the sensor data is further based at least in part on determining less than a threshold quantity of objects associated with the at least a portion of the sensor data classified as conforming to the user-specified category correspond to the particular type of object.
15. The method of claim 1 , wherein the at least a portion of the sensor data classified as conforming to the user-specified category identifies an object is located in a particular region.
16. The method of claim 1 , wherein the at least a portion of the sensor data classified as conforming to the user-specified category identifies sensor data associated with a performance of a particular machine learning model less than a threshold value.
17. The method of claim 1 , wherein the sensor data is captured during a plurality of sensor data capture sessions, wherein each of the plurality of sensor data capture sessions correspond to a sequence of sensor data associated with a vehicle trip.
18. The method of claim 1 , wherein the sensor data is captured during a plurality of sensor data capture sessions, and wherein the sensor data comprises first sensor data captured by a first sensor associated with a vehicle during the plurality of sensor data capture sessions and second sensor data captured by a second sensor associated with the vehicle during the plurality of sensor data capture sessions.
19. A system, comprising:
at least one processor, and
at least one non-transitory storage media storing instructions that, when executed by the at least one processor, cause the at least one processor to:
receive sensor data captured by at least one sensor;
process the sensor data using a plurality of machine learning models to identify a plurality of capture session collections represented in the sensor data, a machine learning model of the plurality of machine learning models configured to identify at least a portion of the sensor data as conforming to a category of capture session from a plurality of categories of capture session; and
filter the sensor data based at least partly on a user-specified category of the plurality of categories of capture session to identify a capture session collection, of the plurality of capture session collections, representing at least a portion of the sensor data classified as conforming to the user-specified category.
20. At least one non-transitory storage media storing instructions that, when executed by a computing system comprising a processor, cause the computing system to:
receive sensor data captured by at least one sensor;
process the sensor data using a plurality of machine learning models to identify a plurality of capture session collections represented in the sensor data, a machine learning model of the plurality of machine learning models configured to identify at least a portion of the sensor data as conforming to a category of capture session from a plurality of categories of capture session; and
filter the sensor data based at least partly on a user-specified category of the plurality of categories of capture session to identify a capture session collection, of the plurality of capture session collections, representing at least a portion of the sensor data classified as conforming to the user-specified category.