Using captured video data to identify pose of a vehicle

1. A computer-implemented method of detecting and classifying pose of a vehicle captured in a video sequence, the method comprising:

receiving a video sequence comprising a plurality of digital image frames that contain an image of a vehicle;

generating an image stack by scaling and shifting a set of the digital image frames to a fixed scale and yielding a sequence of images of the vehicle over a time period;

processing the image stack with a classifier to determine a pose of the vehicle as the vehicle appears in the video sequence according to a group of candidate classes that comprise a plurality of the following: REAR, REAR_LEFT, SIDE_LEFT, FRONT_LEFT, FRONT, FRONT_RIGHT, SIDE_RIGHT or REAR-RIGHT;

using the determined pose to predict a direction of movement of the vehicle; and

causing another vehicle to take an action in response to the predicted direction of movement;

wherein predicting the direction of movement comprises: using the determined pose to determine which turn signals are visible on the vehicle; and determining a state for each of the visible turn signals.

2. The method of claim 1 further comprising, before processing the image stack with the classifier, training the classifier on a plurality of training image stack sets that include, for each training image stack, labels indicative of object pose class.

3. The method of claim 1 further comprising, before generating the image stack, cropping each of the digital image frames in the set to eliminate information outside of the bounding boxes of each frame.

4. The method of claim 1 , wherein:

receiving the video sequence is performed by a camera of an autonomous vehicle (AV);

determining the pose of the vehicle is performed by an on-board processor of the AV; and

the method further comprises, by the on-board processor:

using the pose to predict a direction of movement of the vehicle, and

causing the AV to take an action responsive to the predicted direction of movement of the vehicle.

5. The method of claim 1 further comprising, before generating the image stack:

processing the digital image frames to detect the vehicle in the digital image frames by adding bounding boxes to the digital image frames; and

performing registration on the set of digital image frames to cause the bounding boxes of each frame in the set to share in a common location and scale within each digital image frame.

6. The method of claim 5 , wherein processing the digital image frames to detect the vehicle in the digital image frames comprises applying Mask R-CNN to the digital image frames.

7. The method of claim 5 further comprising, before performing the registration, tracking the vehicle across the digital image frames to eliminate frames that are less likely to contain the vehicle, yielding the set on which registration will be performed.

8. The method of claim 7 , wherein tracking the vehicle across the digital image frames comprises:

performing Intersection over Union matching between a plurality of pairs of the digital image frames; or

performing color histogram matching between a plurality of pairs of the digital image frames.

9. An on-board system of a vehicle for detecting and classifying pose of another vehicle observed in

a video sequence, the on-board system comprising:

a video camera;

an on-board processor; and

a computer-readable memory containing programming instructions that are configured to cause the on-board processor to:

receive a video sequence comprising a plurality of digital image frames that contain an image of another vehicle;

generate an image stack by scaling and shifting a set of the digital image frames to a fixed scale and yielding a sequence of images of the other vehicle over a time period;

process the image stack with a classifier to determine a pose of the other vehicle as the other vehicle appears in the video sequence according to a set of candidate classes;

predict a direction of movement of the other vehicle based on the pose; and

cause the on-board processor to instruct a vehicle operational system to take an action in response to the predicted direction of movement;

wherein the programming instructions to predict the direction of movement of the other vehicle comprise instructions to: use the determined pose to determine which turn signals are visible on the vehicle; and determine a state for each of the visible turn signals.

10. The system of claim 9 , wherein the candidate classes comprise a plurality of the following: REAR, REAR_LEFT, SIDE_LEFT, FRONT_LEFT, FRONT, FRONT_RIGHT, SIDE_RIGHT or REAR-RIGHT.

11. The system of claim 9 wherein the classifier is a classifier that has been trained on a plurality of training image stack sets that include, for each training image stack, labels indicative of object pose class.

12. The system of claim 9 further comprising additional programming instructions configured to cause the on-board processor to, before generating the image stack, crop each of the digital image frames in the set to eliminate information outside of the bounding boxes of each frame.

13. The system of claim 9 further comprising additional programming instructions configured to cause the on-board processor to, before generating the image stack:

process the digital image frames to detect the other vehicle in the digital image frames by adding bounding boxes to the digital image frames; and

perform registration on the set of digital image frames to cause the bounding boxes of each frame in the set to share in a common location and scale within each digital image frame.

14. The system of claim 13 , wherein the programming instructions to process the digital image frames to detect the object in the digital image frames comprise instructions to apply Mask R-CNN to the digital image frames.

15. The system of claim 9 further comprising programming instructions configured to cause the on-board processor to, before performing registration, track the other vehicle across the digital image frames to eliminate frames that are less likely to contain the other vehicle, yielding the set on which registration will be performed.

16. The system of claim 15 , wherein the programming instructions to track the other vehicle across the digital image frames comprise instructions to:

perform Intersection over Union matching between a plurality of pairs of the digital image frames; or

perform color histogram matching between a plurality of pairs of the digital image frames.

17. A non-transitory computer-readable medium that stores instructions that is configured to, when executed by at least one computing device, cause the at least one computing device to perform operations comprising:

receiving a video sequence comprising a plurality of digital image frames that contain an image of a vehicle;

generating an image stack by scaling and shifting a set of the digital image frames to a fixed scale and yielding a sequence of images of the vehicle over a time period;

processing the image stack with a classifier to determine a pose of the vehicle as the vehicle appears in the video sequence according to a group of candidate classes that comprise a plurality of the following: REAR, REAR_LEFT, SIDE_LEFT, FRONT_LEFT, FRONT, FRONT_RIGHT, SIDE_RIGHT or REAR-RIGHT;

using the determined pose to predict a direction of movement of the vehicle; and

causing another vehicle to take an action in response to the predicted direction of movement;

wherein predicting the direction of movement comprises: using the determined pose to determine which turn signals are visible on the vehicle; and determining a state for each of the visible turn signals.