SHARING CLASSIFIED OBJECTS PERCEIVED BY AUTONOMOUS VEHICLES
Embodiments are disclosed for sharing classified objects perceived by autonomous vehicles. In an embodiment, a method comprises: obtaining, from an autonomous vehicle (AV), a first scene description, the first scene description including one or more classified objects detected in a first zone of a geographic area; updating, by a first edge node in the first zone and using a plurality of scene descriptions, an autonomous system grid (ASG) for the first zone; and sending, by the first edge node, the updated ASG to a second edge node located in the first zone or in a second zone of the geographic area.
1 . A method comprising:
obtaining, from an autonomous vehicle (AV), a first scene description, the first scene description including one or more classified objects detected in a first zone of a geographic area;
updating, by a first edge node in the first zone and using a plurality of scene descriptions, an autonomous system grid (ASG) for the first zone; and
sending, by the first edge node, the updated ASG to a second edge node located in the first zone or a second zone of the geographic area,
wherein the method is performed by one or more special-purpose computing devices.
2 . The method of claim 1 , further comprising:
detecting, by the first edge node, one or more outlier scene descriptions in the plurality of scene descriptions; and
excluding the one or more outlier scene descriptions from the updating.
3 . The method of claim 1 , further comprising:
determining, by the first edge node, whether a timestamp of the first scene description exceeds an age threshold; and
in accordance with the timestamp exceeding the age threshold,
excluding the first scene description from the updating.
4 . The method of claim 3 , wherein the timestamp of the first scene description is an oldest object timestamp for an object in the first scene description.
5 . The method of claim 1 , further comprising:
generating, by the first edge node, a second scene description generated from data provided by one or more non-AV sensors in the first zone; and
using, by the first edge node, the first and second scene descriptions in the updating.
6 . The method of claim 1 , further comprising:
determining, by the first edge node, a confidence score indicating an accuracy of the ASG.
7 . The method of claim 1 , further comprising:
obtaining, by the first edge node, one or more ASG updates from a cloud based computer platform; and
using, by the first edge node, the one or more ASG updates in the updating.
8 . The method of claim 1 , wherein the updating includes adding or deleting one or more objects to or from the ASG.
9 . The method of claim 1 , wherein updating includes adding or changing labels for one or more objects of the ASG.
10 . The method of claim 1 , wherein updating includes updating at least one state of at least one dynamic object in the ASG.
11 . A system comprising:
a core network;
a plurality of edge nodes coupled to the core network, a first edge node of the plurality of edge nodes including one or more computer processors programmed to:
obtain, from a first autonomous vehicle (AV), a first scene description, the first scene description including one or more classified objects detected in a first zone that includes a first portion of a geographic area;
update an autonomous system grid (ASG) for the first zone using a plurality of scene descriptions; and
send the updated ASG to a second edge node located in the first zone or in a second zone including a second portion of the geographic area.
12 . The system of claim 11 , wherein the one or more computer processors are programmed to:
detect one or more outlier scene descriptions in the plurality of scene descriptions; and
exclude the one or more outlier scene descriptions from the updating.
13 . The system of claim 11 , wherein the one or more computer processors are further programmed to:
determine whether a timestamp of the first scene description exceeds an age threshold; and
in accordance with the timestamp exceeding the age threshold,
exclude the first scene description from the updating.
14 . The system of claim 13 , wherein the timestamp of the first scene description is an oldest object timestamp for an object in the first scene description.
15 . The system of claim 11 , wherein the one or more computer processors are further programmed to:
generate a second scene description from data provided by one or more non-AV sensors in the first zone; and
use the first and second scene descriptions in the updating.
16 . The system of claim 11 , wherein the one or more computer processors are further programmed to:
determine, by the first edge node, a confidence score indicating an accuracy of the ASG.
17 . The system of claim 11 , wherein the one or more computer processors are further programmed to:
obtain, by the first edge node, one or more ASG updates from a cloud based computer platform; and
use, by the first edge node, the one or more ASG updates in the updating.
18 . The system of claim 11 , wherein the updating includes adding or deleting one or more objects to or from the ASG.
19 . The system of claim 11 , wherein updating includes adding or changing labels for one or more objects of the ASG.
20 . The system of claim 11 , wherein updating includes updating at least one state of at least one dynamic object in the ASG.