Navigation based on partially occluded pedestrians

1. A navigation system for a host vehicle, the navigation system comprising:

at least one processor comprising circuitry and a memory, wherein the memory includes instructions that when executed by the circuitry cause the at least one processor to:

receive a captured image acquired by a camera onboard the host vehicle, the captured image including a representation of an occluded pedestrian in an environment of the host vehicle, wherein the captured image lacks a representation of a region where the occluded pedestrian contacts a ground surface, the region where the occluded pedestrian contacts the ground surface being obscured from view in the captured image by at least a portion of the host vehicle;

provide the captured image to an analysis module configured to generate an output for the captured image, wherein the generated output includes an indicator of a contact position of the occluded pedestrian with the ground surface, and wherein the analysis module includes at least one trained model trained based on training data including a plurality of training images, each of the plurality of images including:

a representation of at least one training pedestrian, and

a region where the training pedestrian contacts a training ground surface,

and having been modified to occlude the region where the training pedestrian contacts the training ground surface;

receive from the analysis module the generated output, including the indicator of the contact position of the occluded pedestrian with the ground surface, the indicator of the contact position of the occluded pedestrian with the ground surface including a bounding box, the bounding box including an area of the captured image including a representation of the at least a portion of the host vehicle; and

cause at least one navigational action by the host vehicle based on the indicator of the contact position of the occluded pedestrian with the ground surface.

2. The navigation system of claim 1 , wherein the training data further includes LIDAR depth information corresponding to the plurality of training images.

3. The navigation system of claim 1 , wherein the at least one trained model includes a neural network.

4. The navigation system of claim 1 , wherein the bounding box is used in a determination of an estimated time to collision between the host vehicle and the pedestrian.

5. The navigation system of claim 1 , wherein the bounding box is used in a determination of an estimated distance between the pedestrian and at least a portion of the host vehicle.

6. The navigation system of claim 1 , wherein the bounding box surrounds the representation of the at least one occluded pedestrian in the captured image and is sized to surround the region where the occluded pedestrian contacts the ground surface had such a region been represented in the captured image.

7. The navigation system of claim 1 , wherein the bounding box is defined by at least two pixel dimensions relative to a reference pixel associated with the captured image.

8. The navigation system of claim 1 , wherein the bounding box is defined by at least two real world dimensions.

9. The navigation system of claim 1 , wherein the at least a portion of the host vehicle includes at least a portion of a hood of the host vehicle.

10. The navigation system of claim 1 , wherein the ground surface includes at least one of a road surface or a sidewalk surface.

11. The navigation system of claim 1 , wherein the at least one navigational action includes at least one of accelerating, braking, or turning the host vehicle.

12. The navigation system of claim 1 , wherein the memory further includes instructions that when executed by the circuitry cause the at least one processor to:

maintain a buffer including a plurality of buffer images acquired prior to the captured image, wherein one or more of the plurality of buffer images include a representation of the occluded pedestrian of the captured image.

13. The navigation system of claim 12 , wherein one or more of the plurality of buffer images include a representation of the region where the occluded pedestrian contacts the ground surface.

14. The navigation system of claim 12 , wherein the plurality of buffer images includes up to 20 acquired image frames.

15. The navigation system of claim 12 ,

wherein the memory further includes instructions that when executed by the circuitry cause the at least one processor to provide one or more of the plurality of buffer images to the analysis module, and

wherein the indicator of the contact position of the occluded pedestrian with the ground surface is determined, at least in part, based on the one or more of the plurality of buffer images.

16. A method for navigating a host vehicle, the method comprising:

receiving a captured image acquired by a camera onboard the host vehicle, the captured image including a representation of an occluded pedestrian in an environment of the host vehicle, wherein the captured image lacks a representation of a region where the occluded pedestrian contacts a ground surface, the region where the occluded pedestrian contacts the ground surface being obscured from view in the captured image by at least a portion of the host vehicle;

providing the captured image to an analysis module configured to generate an output for the captured image, wherein the generated output includes an indicator of a contact position of the occluded pedestrian with the ground surface, and wherein the analysis module includes at least one trained model trained based on training data including a plurality of training images, each of the plurality of images including:

a representation of at least one training pedestrian, and

a region where the training pedestrian contacts a training ground surface,

and having been modified to occlude the region where the training pedestrian contacts the training ground surface;

receiving from the analysis module the generated output, including the indicator of the contact position of the occluded pedestrian with the ground surface, the indicator of the contact position of the occluded pedestrian with the ground surface including a bounding box, the bounding box including an area of the captured image including a representation of the at least a portion of the host vehicle; and

causing at least one navigational action by the host vehicle based on the indicator of the contact position of the occluded pedestrian with the ground surface.

17. A non-transitory computer-readable medium storing instructions that, when executed by at least one processor, are configured to cause at least one processor to perform a method for navigating a host vehicle, the method comprising:

receiving a captured image acquired by a camera onboard the host vehicle, the captured image including a representation of an occluded pedestrian in an environment of the host vehicle, wherein the captured image lacks a representation of a region where the occluded pedestrian contacts a ground surface, the region where the occluded pedestrian contacts the ground surface being obscured from view in the captured image by at least a portion of the host vehicle;

providing the captured image to an analysis module configured to generate an output for the captured image, wherein the generated output includes an indicator of a contact position of the occluded pedestrian with the ground surface, and wherein the analysis module includes at least one trained model trained based on training data including a plurality of training images, each of the plurality of images including:

a representation of at least one training pedestrian, and

a region where the training pedestrian contacts a training ground surface,

and having been modified to occlude the region where the training pedestrian contacts the training ground surface;

receiving from the analysis module the generated output, including the indicator of the contact position of the occluded pedestrian with the ground surface, the indicator of the contact position of the occluded pedestrian with the ground surface including a bounding box, the bounding box including an area of the captured image including a representation of the at least a portion of the host vehicle; and

causing at least one navigational action by the host vehicle based on the indicator of the contact position of the occluded pedestrian with the ground surface.