System and method for predicting driver situational awareness
A system and method for predicting a driver's situational awareness that includes receiving driving scene data associated with a driving scene of an ego vehicle and eye gaze data to track a driver's eye gaze behavior with respect to the driving scene. The system and method also include analyzing the eye gaze data and determining an eye gaze fixation value associated with each object that is located within the driving scene and analyzing the driving scene data and determining a situational awareness probability value associated with each object that is located within the driving scene that is based on a salience, effort, expectancy, and a cost value associated with each of the objects within the driving scene. The system and method further include communicating control signals to electronically control at least one component based on the situational awareness probability value and the eye gaze fixation value.
1 . A computer-implemented method for predicting a driver's situational awareness comprising:
receiving driving scene data associated with a driving scene of an ego vehicle and eye gaze data to track a driver's eye gaze behavior with respect to the driving scene;
analyzing the eye gaze data and determining an eye gaze fixation value for each and every object that is located within the driving scene based on an eye gaze fixation time for each of the objects within the driving scene, wherein determining the eye gaze fixation value includes establishing a global range of eye gaze fixation times for all objects within the driving scene, the global range comprising a global minimum and maximum fixation time across the objects present in the driving scene, computing the eye gaze fixation value for each of the objects within the driving scene by scaling that object's eye gaze fixation time with respect to the global range, and assigning a zero eye gaze fixation value to each object in the driving scene that is not intersected by the driver's gaze, thereby ensuring that all objects are represented in the comparative analysis;
analyzing the driving scene data and determining a situational awareness probability value associated with each object that is located within the driving scene that is based on a salience, effort, expectancy, and a cost value associated with each of the objects within the driving scene; and
communicating control signals to electronically control at least one component of the ego vehicle based on the situational awareness probability value and the eye gaze fixation value associated with each of the objects that are located within the driving scene.
2 . The computer-implemented method of claim 1 , wherein analyzing the eye gaze data and determining the eye gaze fixation value includes extracting eye gaze coordinates from the eye gaze data and analyzing the driving scene data and the eye gaze coordinates that are associated with matching time steps to identify a correspondence between various portions of the driving scene and the driver's eye gaze behavior.
3 . The computer-implemented method of claim 2 , wherein analyzing the eye gaze data and determining the eye gaze fixation value includes analyzing time coordinates associated with the eye gaze fixation time of the driver's eye gaze behavior with respect to different portions of the driving scene and labeling each object with the eye gaze fixation time of the driver's eye gaze behavior that are included within respective different portions of the driving scene, wherein the eye gaze fixation value is determined as a numerical value that pertains to a labeled eye gaze fixation time with respect to each object, and
the eye gaze fixation value and the situational awareness probability value are aggregated to process a driver object awareness value which pertains to the driver's eye gaze fixation time and a predicted level of situational awareness that the driver of the ego vehicle has with respect to each of the objects that are located within the driving scene.
4 . The computer-implemented method of claim 1 , wherein analyzing the driving scene data and determining the situational awareness probability value includes analyzing the driving scene data at a respective timeframe at which eye gaze fixations of the driver have been captured to extract object property-based features associated with each object located within the driving scene.
5 . The computer-implemented method of claim 4 , wherein the object property-based features associated with each object located within the driving scene include at least one of: object contrast, object movement, object relevance, object priority, object size, and object proximity.
6 . The computer-implemented method of claim 5 , wherein determining the situational awareness probability value associated with each object that is located within the driving scene includes determining a salience level that is related to the salience associated with each of the objects based on an analysis of the object property-based features, wherein the salience level pertains to at least one of: a likelihood that each object captures an attention of the driver of the ego vehicle and a product of object contrast of each object against a background of the driving scene.
7 . The computer-implemented method of claim 5 , wherein determining the situational awareness probability value associated with each object that is located within the driving scene includes determining an expectancy level that is related to the expectancy associated with each of the objects based on an analysis of the object property-based features, wherein the expectancy level pertains to at least one of: a respective position of each object as compared to a respective position of the ego vehicle and a product of an eye gaze fixation of the driver of the ego vehicle at a particular point in time with respect to each object.
8 . The computer-implemented method of claim 5 , wherein determining the situational awareness probability value associated with each object that is located within the driving scene includes determining an effort level that is related to the effort associated with each of the objects based on an analysis of the object property-based features, wherein the effort level pertains to at least one of: a level of effort that is involved with respect to the driver of the ego vehicle attending to various objects that are located within the driving scene and an object density of the driving scene.
9 . The computer-implemented method of claim 5 , wherein determining the situational awareness probability value associated with each object that is located within the driving scene includes determining the cost value associated with each of the objects based on an analysis of the object property-based features, wherein the cost value is representative of cost of missing information associated with each object that is based on at least one of: a classification of each object and a potential likelihood of overlap between a projected path of each respective object and a projected path of the ego vehicle.
10 . A system for predicting a driver's situational awareness comprising:
a memory storing instructions when executed by a processor cause the processor to:
receive driving scene data associated with a driving scene of an ego vehicle and eye gaze data to track a driver's eye gaze behavior with respect to the driving scene;
analyze the eye gaze data and determining an eye gaze fixation value for each and every object that is located within the driving scene based on an eye gaze fixation time for each of the objects within the driving scene, wherein determining the eye gaze fixation value includes establishing a global range of eye gaze fixation times for all objects within the driving scene, the global range comprising a global minimum and maximum fixation time across the objects present in the driving scene, computing the eye gaze fixation value for each of the objects within the driving scene by scaling that object's eye gaze fixation time with respect to the global range, and assigning a zero eye gaze fixation value to each object in the driving scene that is not intersected by the driver's gaze, thereby ensuring that all objects are represented in the comparative analysis;
analyze the driving scene data and determining a situational awareness probability value associated with each object that is located within the driving scene that is based on a salience, effort, expectancy, and a cost value associated with each of the objects within the driving scene; and
communicate control signals to electronically control at least one component of the ego vehicle based on the situational awareness probability value and the eye gaze fixation value associated with each of the objects that are located within the driving scene.
11 . The system of claim 10 , wherein analyzing the eye gaze data and determining the eye gaze fixation value includes extracting eye gaze coordinates from the eye gaze data and analyzing the driving scene data and the eye gaze coordinates that are associated with matching time steps to identify a correspondence between various portions of the driving scene and the driver's eye gaze behavior.
12 . The system of claim 11 , wherein analyzing the eye gaze data and determining the eye gaze fixation value includes analyzing time coordinates associated with the eye gaze fixation time of the driver's eye gaze behavior with respect to different portions of the driving scene and labeling each object with the eye gaze fixation time of the driver's eye gaze behavior that are included within respective different portions of the driving scene, wherein the eye gaze fixation value is determined as a numerical value that pertains to a labeled eye gaze fixation time with respect to each object, and
the eye gaze fixation value and the situational awareness probability value are aggregated to process a driver object awareness value which pertains to the driver's eye gaze fixation time and a predicted level of situational awareness that the driver of the ego vehicle has with respect to each of the objects that are located within the driving scene.
13 . The system of claim 10 , wherein analyzing the driving scene data and determining the situational awareness probability value includes analyzing the driving scene data at a respective timeframe at which eye gaze fixations of the driver have been captured to extract object property-based features associated with each object located within the driving scene.
14 . The system of claim 13 , wherein the object property-based features associated with each object located within the driving scene include at least one of: object contrast, object movement, object relevance, object priority, object size, and object proximity.
15 . The system of claim 14 , wherein determining the situational awareness probability value associated with each object that is located within the driving scene includes determining a salience level that is related to the salience associated with each of the objects based on an analysis of the object property-based features, wherein the salience level pertains to at least one of: a likelihood that each object captures an attention of the driver of the ego vehicle and a product of object contrast of each object against a background of the driving scene.
16 . The system of claim 14 , wherein determining the situational awareness probability value associated with each object that is located within the driving scene includes determining an expectancy level that is related to the expectancy associated with each of the objects based on an analysis of the object property-based features, wherein the expectancy level pertains to at least one of: a respective position of each object as compared to a respective position of the ego vehicle and a product of an eye gaze fixation of the driver of the ego vehicle at a particular point in time with respect to each object.
17 . The system of claim 14 , wherein determining the situational awareness probability value associated with each object that is located within the driving scene includes determining an effort level that is related to the effort associated with each of the objects based on an analysis of the object property-based features, wherein the effort level pertains to at least one of: a level of effort that is involved with respect to the driver of the ego vehicle attending to various objects that are located within the driving scene and an object density of the driving scene.
18 . The system of claim 14 , wherein determining the situational awareness probability value associated with each object that is located within the driving scene includes determining the cost value associated with each of the objects based on an analysis of the object property-based features, wherein the cost value is representative of cost of missing information associated with each object that is based on at least one of: a classification of each object and a potential likelihood of overlap between a projected path of each respective object and a projected path of the ego vehicle.
19 . A non-transitory computer readable storage medium storing instructions that when executed by a computer, which includes a processor performs a method, the method comprising:
receiving driving scene data associated with a driving scene of an ego vehicle and eye gaze data to track a driver's eye gaze behavior with respect to the driving scene;
analyzing the eye gaze data and determining an eye gaze fixation value for each and every object that is located within the driving scene that is based on an eye gaze fixation time for each of the objects within the driving scene, wherein determining the eye gaze fixation value includes establishing a global range of eye gaze fixation times for all objects within the driving scene, the global range comprising a global minimum and maximum fixation time across the objects present in the driving scene, computing the eye gaze fixation value for each of the objects within the driving scene by scaling that object's eye gaze fixation time with respect to the global range, and assigning a zero eye gaze fixation value to each object in the driving scene that is not intersected by the driver's gaze, thereby ensuring that all objects are represented in the comparative analysis;
analyzing the driving scene data and determining a situational awareness probability value associated with each object that is located within the driving scene that is based on a salience, effort, expectancy, and a cost value associated with each of the objects within the driving scene; and
communicating control signals to electronically control at least one component of the ego vehicle based on the situational awareness probability value and the eye gaze fixation value associated with each of the objects that are located within the driving scene.
20 . The non-transitory computer readable storage medium of claim 19 , wherein analyzing the driving scene data and determining the situational awareness probability value includes analyzing the driving scene data at a respective timeframe at which eye gaze fixations of the driver have been captured to extract object property-based features associated with each object located within the driving scene, wherein the object property-based features associated with each object located within the driving scene include at least one of: object contrast, object movement, object relevance, object priority, object size, and object proximity.