VISUAL LOGIN
Systems and methods are provided for intelligent driving monitoring systems, advanced driver assistance systems and autonomous driving systems, and providing one-click driver image labeling to facilitate the assignment of drive time to a driver. Various aspects of the driver and/or passengers may be determined, which may include a driver identifier previously associated with the driver, whether the driver is wearing a facial covering, whether an image of the driver is partially occluded, whether the driver is facing in the direction of the camera, and whether an image of the driver is expected to be easily recognizable to a person tasked with assigning representative driver images to driver identifiers. These determinations may be used to improve the efficacy and ease-of-use of a visual login system.
1 . A method for monitoring a driver of a vehicle, the method comprising:
monitoring, by at least one processor coupled to a driver-facing camera, a face of the driver, wherein the driver-facing camera is arranged to have a view of an interior of the vehicle;
detecting, by the at least one processor, an obstruction between the driver-facing camera and the face of the driver based on an image from the driver-facing camera; and
responsive to detecting the obstruction in the image, delaying, by the at least one processor, processing of the driver using the image.
2 . The method of claim 1 , wherein the obstruction is glasses worn by the driver.
3 . The method of claim 1 , wherein the obstruction is a baseball cap worn by the driver.
4 . The method of claim 1 , wherein the obstruction is facial hair.
5 . The method of claim 1 , wherein the obstruction is a sun visor.
6 . The method of claim 1 , wherein the obstruction is a dangling wire situated between the camera and the driver.
7 . The method of claim 1 , further comprising:
preparing training data for a visual login system, wherein the training data comprises a plurality of images of the driver.
8 . The method of claim 7 , wherein preparing training data comprises augmenting the training data with the image corresponding to the detected obstruction between the driver-facing camera and the face of the driver.
9 . The method of claim 7 , wherein delaying processing of the driver comprises selecting a representative image of the driver for which the face of the driver is not obstructed.
10 . The method of claim 1 , further comprising disabling the camera in response to detecting the obstruction.
11 . The method of claim 1 , further comprising:
detecting that the driver was distracted or drowsy for an unsafe period of time based at least in part on the image; wherein delaying processing of the driver using the image comprises classifying the detection that the driver was distracted or drowsy as a mistake.
12 . The method of claim 1 , further comprising:
detecting that the driver was distracted or drowsy for an unsafe period of time based at least in part on the image; and
preparing training data for a driver monitoring system, wherein the training data comprises the image of the driver, wherein the image corresponds to the detection that the driver was distracted or drowsy and the detection of the obstruction.
13 . A driver monitoring system, comprising:
a driver-facing camera having a view of an interior of a vehicle; and
a processing device configured to:
detect an obstruction between the camera and the face of a driver of the vehicle based on an image from the driver-facing camera; and
delay processing of the driver in response to detecting the obstruction in the image.
14 . The driver monitoring system of claim 13 , wherein the obstruction is one of:
glasses worn by the driver;
a baseball cap worn by the driver;
facial hair;
a sun visor; or
a dangling wire situated between the camera and the driver.
15 . The driver monitoring system of claim 13 , wherein the processing device is further configured to:
prepare training data for a visual login system, wherein the training data comprises a plurality of images of the driver.
16 . The driver monitoring system of claim 15 , wherein preparing training data comprises augmenting the training data with the image corresponding to the detected obstruction between the driver-facing camera and the face of the driver.
17 . The driver monitoring system of claim 15 , wherein delaying processing of the driver comprises selecting a representative image of the driver for which the face of the driver is not obstructed.
18 . The driver monitoring system of claim 13 wherein the processing device is configured to disable the camera in response to detecting the obstruction.
19 . The driver monitoring system of claim 13 , wherein the processing device is further configured to:
detect that the driver was distracted or drowsy for an unsafe period of time;
wherein delaying processing of the driver comprises classifying the detection that the driver was distracted or drowsy as a mistake.
20 . The driver monitoring system of claim 13 , wherein the processing device is further configured to:
prepare training data for the driver monitoring system, wherein the training data comprises the image of the driver, wherein the image corresponds to the detection that the driver was distracted and the detection of the obstruction.