Gesture based authentication for autonomous vehicles

1. A method of authenticating a user of an autonomous vehicle (AV) comprising:

receiving a selection of an authentication gesture associated with a user, the authentication gesture selected from a plurality of authentication gestures, the authentication gesture comprising at least two different body positions;

receiving sensor data describing an environment surrounding the AV, the sensor data comprising data from at least one light detecting and ranging (lidar) sensor;

identifying, based on the sensor data, the user in the environment of the AV;

comparing sensor data collected over a period of time describing a series of body positions of the identified user to a model of a human performing the authentication gesture, the model trained based on sensor observations of a plurality of humans performing the authentication gesture;

determining, based on the comparison, that the identified user is performing the authentication gesture; and

in response to determining that the identified user is performing the authentication gesture, authorizing the identified user to access the AV.

2. The method of claim 1 , wherein the sensor data further comprises data from at least one of a camera and a radar sensor.

3. The method of claim 1 , wherein the model of a human performing the authentication gesture comprises a plurality of models of limbs in a plurality of limb positions, and comparing sensor data collected over a period of time describing the series of body positions of the identified user to a model of a human performing the authentication gesture comprises:

identifying a limb of the identified user in the sensor data; and

comparing data describing a position of the identified limb to one of the plurality of models of limbs.

4. The method of claim 1 , further comprising:

selecting, from a gesture database storing the plurality of authentication gestures, the authentication gesture; and

transmitting, to a user device associated with the user, instructions to perform the authentication gesture to obtain authorization to access the AV.

5. The method of claim 4 , wherein the instructions comprise instructions for the user to stand within a range of positions relative to the AV, and identifying the user in the environment of the AV comprises:

analyzing sensor data corresponding to the range of positions relative to the AV, and

identifying the user located within the range of positions relative to the AV.

6. The method of claim 1 , further comprising:

in response to identifying the user in the environment of the AV, providing an instruction to the identified user to perform the authentication gesture; and

after providing the instruction to the identified user to perform the gesture, collecting the data describing the series of body positions of the identified user.

7. The method of claim 1 , further comprising outputting, from the AV, at least one of an auditory signal and a visual signal indicating that the AV has authorized the identified user to access the AV.

8. The method of claim 1 , further comprising:

identifying, based on the sensor data, a plurality of users in the environment of the AV, the plurality of users including the identified user;

determining a number of the plurality of users; and

authorizing the identified user to access the AV further in response to the number of the plurality of users being equal to a number of expected users.

9. The method of claim 1 , further comprising:

determining, based on the sensor data, at least one physical characteristic of the identified user;

comparing the at least one physical characteristic of the identified user to at least one physical characteristic of an intended user of the AV; and

authorizing the identified user to access the AV further in response to the at least one physical characteristic of the identified user matching the at least one physical characteristic of the intended user.

10. The method of claim 1 , wherein the authentication gesture is selected from the plurality of authentication gestures by the user.

11. The method of claim 1 , wherein a second user is associated with a second authentication gesture of the plurality of authentication gestures, the second authentication gesture to authenticate the second user to a second AV.

12. A system for authenticating a user of an autonomous vehicle (AV) comprising:

a light detecting and ranging (lidar) sensor configured to collect data describing an environment surrounding the AV; and

a processor configured to:

receive a selection of an authentication gesture associated with a user, the authentication gesture selected from a plurality of authentication gestures, the authentication gesture comprising at least two different body positions;

identify, based on the collected data, the user in the environment of the AV;

select a portion of the collected data, the selected portion describing a series of body positions of the identified user;

compare the selected portion of the collected data describing the series of body positions of the identified user to a model of a human performing the authentication gesture, the model trained based on sensor observations of a plurality of humans performing the authentication gesture;

determine, based on the comparison, that the identified user is performing the authentication gesture; and

in response to determining that the identified user is performing the authentication gesture, authorize the identified user to access the AV.

13. The system of claim 12 , further comprising at least one of:

a camera configured to collect images of the environment surrounding the AV; and

a radar sensor configured to collect data describing the environment surrounding the AV.

14. The system of claim 12 , wherein the model of a human performing the authentication gesture comprises a plurality of models of limbs in a plurality of limb positions, and the processor is configured to compare the selected portion of the collected data to the model of a human performing a gesture by:

identifying a limb of the identified user in the collected data; and

comparing data describing a position of the identified limb to one of the plurality of models of limbs.

15. The system of claim 12 , wherein the processor is further configured to:

analyze sensor data corresponding to a range of positions relative to the AV, the range of positions provided to the user in an instruction to perform the authentication gesture, and

identify the user located within the range of positions relative to the AV.

16. The system of claim 12 , wherein the processor is further configured to:

in response to identifying the user in the environment of the AV, provide an instruction to the user to perform the authentication gesture; and

after providing the instruction to the user to perform the authentication gesture, collect the data describing the series of body positions of the identified user.

17. The system of claim 12 , wherein the processor is further configured to:

identify, based on the collected data, a plurality of users in the environment of the AV, the plurality of users including the identified user;

determine a number of the plurality of users; and

authorize the identified user to access the AV further in response to the number of the plurality of users being equal to a number of expected users.

18. The system of claim 12 , wherein the processor is further configured to:

determine, based on the collected data, at least one physical characteristic of the identified user;

compare the at least one physical characteristic of the identified user to at least one physical characteristic of an intended user of the AV; and

authorize the identified user to access the AV further in response to the at least one physical characteristic of the identified user matching the at least one physical characteristic of the intended user.

19. A system for authenticating a user of an autonomous vehicle (AV) comprising:

a sensor suite configured to collect data describing an environment surrounding the AV, the sensor suite comprising a light detecting and ranging (lidar) sensor, a camera, and a radar sensor; and

a processor configured to:

receive a selection of an authentication gesture associated with a user, the authentication gesture selected from a plurality of authentication gestures, the authentication gesture comprising at least two different body positions;

identify, based on the collected data, the user in the environment of the AV;

compare sensor data collected over a period of time describing a series of body positions of the identified user to a model of a human performing the authentication gesture, the model trained based on sensor observations of a plurality of humans performing the authentication gesture;

determine, based on the comparison, that the identified user is performing the authentication gesture; and

in response to determining that the identified user is performing the authentication gesture, authorize the identified user to access the AV.

20. The system of claim 19 , wherein identifying the user in the environment of the AV comprises comparing the collected data to a model of a human, the model trained based on observations by lidar sensors, cameras, and radar sensors of a plurality of humans in a plurality of environments of a plurality of AVs.