Security operations of parked vehicles
Systems, methods and apparatus of vehicle security operations during parking. For example, a vehicle includes: a proximity sensor configured to detect presence of an object approaching the vehicle when the vehicle is in a parking state; at least one camera configured to monitor surroundings of the vehicle when the vehicle is in the parking state; and an artificial neural network configured to extract identification information of the object from images generated by the camera and determine a security classification of the presence of the object. The identification information is stored in the vehicle and/or transmitted to a server or a mobile device, in response to the security classification being in a predefined category.
1 . A device, comprising:
a camera configured on a vehicle to generate images of surroundings of the vehicle; and
a computing device configured to perform computations on an artificial neural network to identify an object from the images and determine a classification of the object, wherein the artificial neural network includes a spiking neural network;
wherein the computing device is configured to, in response to the classification determined for the object being in a predefined category, record a video clip of the object;
wherein, in response to the classification being in the predefined category, the vehicle is configured to record the video clip of the object for a duration in which the presence of the object is detected by a proximity sensor;
wherein the artificial neural network is configured to be trained, in the vehicle, to predict a user identified security classification based on a video clip being assigned the user identified security classification, wherein the security classification identifies an abnormal activity; the vehicle is configured to record a video clip of the abnormal activity; and after the video clip is reviewed to identify a detailed classification, the artificial neural network is further trained to predict the detailed classification from the video clip.
2 . The device of claim 1 , wherein the object includes an image of a face when the artificial neural network classifies the object as person.
3 . The device of claim 2 , wherein the artificial neural network is configured to recognize a license plate number from the image of the license plate.
4 . The device of claim 3 , further comprising a proximity sensor configured to cause the activation of the camera to generate images of the surroundings of the vehicle, in response to a detection of the presence of the object.
5 . The device of claim 4 , further comprising:
an advanced driver assistance system configured to receive, when in an autonomous driving mode, images from the at least one camera to generate vehicle controls for autonomous driving of the vehicle.
6 . The device of claim 1 , wherein the object includes an image of a license plate when the artificial neural network classifies the object as vehicle.
7 . The device of claim 1 , wherein an advanced driver assistance system is configured to receive at least a portion of outputs of the artificial neural network.
8 . The device of claim 1 , wherein the vehicle is configured to transmit the identification information to a server or a mobile device in response to a security classification.
9 . The device of claim 1 , wherein the predefined category is associated with a potential accident, a potential collision, a break-in attempt, or an attack on the vehicle, or any combination thereof.
10 . The device of claim 1 , further comprising:
an infotainment system configured to generate a warning or alert signal in response to the classification.
11 . The device of claim 1 , further comprising:
at least one sensor configured to generate signals to detect an instance of burglary or vandalism;
wherein the artificial neural network is configured to generate the classification further based on the signals generated by the at least one sensor.
12 . The device of claim 11 , wherein the at least one sensor includes a touch sensor, an impact sensor, a break-in sensor, or a microphone, or any combination thereof.
13 . A method, comprising:
monitoring, by at least one camera configured on a vehicle, surroundings of the vehicle;
providing, to an artificial neural network, images generated by the at least one camera, wherein the artificial neural network includes a spiking neural network;
extracting, by the artificial neural network, identification information of the object from the images generated by the at least one camera;
determining, by the artificial neural network, a classification in relation with the presence of the object in vicinity of the vehicle, wherein the classification identifies an abnormal activity;
storing, in a data storage device of the vehicle, the identification information in response to the classification being in a predefined category;
training, in the vehicle, the artificial neural network to predict a user identified classification based on a video clip being assigned the user identified classification;
recording, by the vehicle, a video clip of the abnormal activity; and
training, in the vehicle, the artificial neural network to predict a user identified security classification based on a video clip being assigned the user identified security classification, wherein the security classification identifies an abnormal activity; the vehicle is configured to record a video clip of the abnormal activity; and after the video clip is reviewed to identify a detailed classification, the artificial neural network is further trained to predict the detailed classification from the video clip.
14 . The method of claim 13 , wherein the identification information is not stored when the classification is not in the predefined category; and the method further comprises:
activating the at least one camera to generate the images in response to the object being detected by a proximity sensor.
15 . The method of claim 13 , further comprising:
recording, into a data storage device of the vehicle, a video clip of the object being in vicinity of the vehicle as detected by a proximity sensor, in response to the classification.
16 . A non-transitory computer storage medium storing instructions which, when executed in a computing device, causes the computing device to perform a method, the method comprising:
monitoring, by at least one camera configured on a vehicle, surroundings of the vehicle;
providing, to an artificial neural network, images generated by the at least one camera, wherein the artificial neural network includes a spiking neural network;
extracting, by the artificial neural network, identification information of the object from the images generated by the at least one camera;
determining, by the artificial neural network, a classification in relation with the presence of the object in vicinity of the vehicle, wherein the classification identifies an abnormal activity;
storing, in a data storage device of the vehicle, the identification information in response to the classification being in a predefined category;
training, in the vehicle, the artificial neural network to predict a user identified classification based on a video clip being assigned the user identified classification;
recording, by the vehicle, a video clip of the abnormal activity; and
training the artificial neural network to predict a user identified security classification based on a video clip being assigned the user identified security classification, wherein the security classification identifies an abnormal activity; the vehicle is configured to record a video clip of the abnormal activity; and after the video clip is reviewed to identify a detailed classification, the artificial neural network is further trained to predict the detailed classification from the video clip.
17 . The non-transitory computer storage medium of claim 16 , wherein the classification is determined from at least images generated by the at least one camera after the extracting of the identification information.