| CPC G06V 20/58 (2022.01) [G06V 10/774 (2022.01); B60W 60/001 (2020.02); B60W 2420/403 (2013.01); B60W 2420/408 (2024.01); B60W 2556/45 (2020.02)] | 20 Claims |
|
1. A method comprising:
receiving, by a processing device, a set of sensor data for a first autonomous vehicle (AV) corresponding to first scene data previously captured for the first AV, wherein the first scene data is part of an intersecting field of view (FoV) with second scene data of a second AV different from the first AV;
inputting the set of sensor data to a simulated trained object detection model of the first AV to generate a first set of object detections for the first scene data, wherein the simulated trained object detection model is an emulated version of an object detection model deployed on the first AV and is executed within a simulation platform using replicated sensor data corresponding to the first scene data;
comparing the first set of object detections with a second set of object detections generated for the second scene data of the second AV by a simulated trained object detection model of the second AV based on sensor data from the second AV;
supplementing, based on the comparing, the first set of object detections with information from the second set of object detections using overlapping tracked objects identified as being detected by both the first AV and the second AV in order to generate a supplemental set of object detections; and
generating, using the supplemental set of object detections, a training data set for an object detection model of the first AV.
|