Abstract: Described herein are systems and methods for determining if a vehicle is parked. In one example, a system includes a processor, a sensor system, and a memory. Both the sensor system and the memory are in communication with the processor. The memory includes a parking determination module having instructions that, when executed by the processor, cause the processor to determine, using a random forest model, when the vehicle is parked based on vehicle estimated features, vehicle learned features, and vehicle taillight features of the vehicle that are based on sensor data from the sensor system.

Abstract: A method includes generating a training data set comprising a plurality of training examples, wherein each training example is generated by receiving map data associated with a road portion, receiving sensor data associated with a road agent located on the road portion, defining one or more corridors associated with the road portion based on the map data and the sensor data, extracting a plurality of agent features associated with the road agent based on the sensor data, extracting a plurality of corridor features associated with each of the one or more corridors based on the sensor data, and for each corridor, labeling the training example based on the position of the road agent with respect to the corridor, and training a neural network using the training data set.

Abstract: An autonomous vehicle can be navigated through an intersection. Topological information about the intersection can be obtained. The topological information can include, for example, a count of a number of lanes on roads associated with the intersection, information about an entrance point to the intersection, or information about an exit point from the intersection. Based on the topological information, context information about a candidate trajectory through the intersection can be obtained. For example, the context information can be based on a current time or information about a position of an object with respect to the topological information. Based on the context information, an existence of an advantage to change an actual trajectory of the autonomous vehicle can be determined. In response to a determination of the existence of the advantage, a change to the actual trajectory of the autonomous vehicle can be caused to occur.

Abstract: A method for traffic light auto-labeling includes aggregating vehicle-to-infrastructure (V2I) traffic light signals at an intersection to determine transition states of each driving lane at the intersection during operation of an ego vehicle. The method also includes automatically labeling image training data to form auto-labeled image training data for a traffic light recognition model within the ego vehicle according to the determined transition states of each driving lane at the intersection. The method further includes planning a trajectory of the ego vehicle to comply with a right-of-way according to the determined transition states of each driving lane at the intersection according to a trained traffic light detection model. A federated learning module may train the traffic light recognition model using the auto-labeled image training data during the operation of the ego vehicle.

Abstract: A method for behavior cloned vehicle trajectory planning is described. The method includes perceiving vehicles proximate an ego vehicle in a driving environment, including a scalar confidence value of each perceived vehicle. The method also includes generating a bird's-eye-view (BEV) grid showing the ego vehicle and each perceived vehicle based on each of the scalar confidence values. The method further includes ignoring at least one of the perceived vehicles when the scalar confidence value of the at least one of the perceived vehicles is less than a predetermined value. The method also includes selecting an ego vehicle trajectory based on a cloned expert vehicle behavior policy according to remaining perceived vehicles.

Abstract: A method for tracking an object performed by an object tracking system includes encoding locations of visible objects in an environment captured in a current frame of a sequence of frames. The method also includes generating a representation of a current state of the environment based on an aggregation of the encoded locations and an encoded location of each object visible in one or more frames of the sequence of frames occurring prior to the current frame. The method further includes predicting a location of an object occluded in the current frame based on a comparison of object centers decoded from the representation of the current state to object centers saved from each prior representation associated with a different respective frame of the sequence of frames occurring prior to the current frame. The method still further includes adjusting a behavior of an autonomous agent in response to identifying the location of the occluded object.

Abstract: Described herein are systems and methods for determining if a vehicle is parked. In one example, a system includes a processor, a sensor system, and a memory. Both the sensor system and the memory are in communication with the processor. The memory includes a parking determination module having instructions that, when executed by the processor, cause the processor to determine, using a random forest model, when the vehicle is parked based on vehicle estimated features, vehicle learned features, and vehicle taillight features of the vehicle that are based on sensor data from the sensor system.

Abstract: A method for tracking an object performed by an object tracking system includes encoding locations of visible objects in an environment captured in a current frame of a sequence of frames. The method also includes generating a representation of a current state of the environment based on an aggregation of the encoded locations and an encoded location of each object visible in one or more frames of the sequence of frames occurring prior to the current frame. The method further includes predicting a location of an object occluded in the current frame based on a comparison of object centers decoded from the representation of the current state to object centers saved from each prior representation associated with a different respective frame of the sequence of frames occurring prior to the current frame. The method still further includes adjusting a behavior of an autonomous agent in response to identifying the location of the occluded object.

Abstract: An action to be performed by a vehicle in response to conflicting input signals can be determined. A sensor signal and an accelerator pedal signal can be received. The sensor signal can indicate a possible need to reduce a speed of the vehicle. The accelerator pedal signal can indicate that an accelerator pedal has been set to a position to increase the speed. An existence of first and second conditions can be determined. The first condition can be that a receipt of the accelerator pedal signal occurred concurrently with a receipt of the sensor signal. The second condition can be that a manner in which the accelerator pedal was set to the position is a better match to a brake pedal usage profile than to an accelerator pedal usage profile. In response to the existence of the first and the second conditions, a brake can be caused to be actuated.

Abstract: A method for simulating, at an instant, the illumination of an indoor scene observed by a camera and being illuminated by outdoor light, the method comprising: a first preliminary phase including: obtaining a reflectance map of the scene elaborating normalized radiosity maps for the contribution of the sky and the ground, a second phase carried out within a first given duration from the instant including: obtaining the position of the sun, obtaining a normalized radiosity map for the contribution of the sun, a third phase, including: acquiring an image of the scene, determining brightness scale parameters for the ground, the sky, and the sun. The disclosure also proposes tracking the camera position.

Abstract: An autonomous vehicle can be navigated through an intersection. Topological information about the intersection can be obtained. The topological information can include, for example, a count of a number of lanes on roads associated with the intersection, information about an entrance point to the intersection, or information about an exit point from the intersection. Based on the topological information, context information about a candidate trajectory through the intersection can be obtained. For example, the context information can be based on a current time or information about a position of an object with respect to the topological information. Based on the context information, an existence of an advantage to change an actual trajectory of the autonomous vehicle can be determined. In response to a determination of the existence of the advantage, a change to the actual trajectory of the autonomous vehicle can be caused to occur.

Abstract: A method for traffic light auto-labeling includes aggregating vehicle-to-infrastructure (V2I) traffic light signals at an intersection to determine transition states of each driving lane at the intersection during operation of an ego vehicle. The method also includes automatically labeling image training data to form auto-labeled image training data for a traffic light recognition model within the ego vehicle according to the determined transition states of each driving lane at the intersection. The method further includes planning a trajectory of the ego vehicle to comply with a right-of-way according to the determined transition states of each driving lane at the intersection according to a trained traffic light detection model. A federated learning module may train the traffic light recognition model using the auto-labeled image training data during the operation of the ego vehicle.

Abstract: An action to be performed by a vehicle in response to conflicting input signals can be determined. A sensor signal and an accelerator pedal signal can be received. The sensor signal can indicate a possible need to reduce a speed of the vehicle. The accelerator pedal signal can indicate that an accelerator pedal has been set to a position to increase the speed. An existence of first and second conditions can be determined. The first condition can be that a receipt of the accelerator pedal signal occurred concurrently with a receipt of the sensor signal. The second condition can be that a manner in which the accelerator pedal was set to the position is a better match to a brake pedal usage profile than to an accelerator pedal usage profile. In response to the existence of the first and the second conditions, a brake can be caused to be actuated.

Abstract: A method for behavior cloned vehicle trajectory planning is described. The method includes perceiving vehicles proximate an ego vehicle in a driving environment, including a scalar confidence value of each perceived vehicle. The method also includes generating a bird's-eye-view (BEV) grid showing the ego vehicle and each perceived vehicle based on each of the scalar confidence value. The method further includes ignoring at least one of the perceived vehicles when the scalar confidence value of the at least one of the perceived vehicles is less than a predetermined value. The method also includes selecting an ego vehicle trajectory based on a cloned expert vehicle behavior policy according to remaining perceived vehicles.

Abstract: A method includes generating a training data set comprising a plurality of training examples, wherein each training example is generated by receiving map data associated with a road portion, receiving sensor data associated with a road agent located on the road portion, defining one or more corridors associated with the road portion based on the map data and the sensor data, extracting a plurality of agent features associated with the road agent based on the sensor data, extracting a plurality of corridor features associated with each of the one or more corridors based on the sensor data, and for each corridor, labeling the training example based on the position of the road agent with respect to the corridor, and training a neural network using the training data set.

Abstract: The disclosure relates to a computer implemented method for building a lighting adaptable map of an indoor scene, the scene comprising at least one light source and at least one surface, wherein the lighting adaptable map adapts its appearance based on a given light setting, wherein the light setting is defined by the state of each light source, comprising: obtaining first image information of the scene, where all light sources are turned on, estimating a map of the scene based on said first image information, said map comprising radiance information and light reflecting characteristics of the surfaces in the scene, detecting and segmenting individual light sources in the scene based on the estimated map, estimating the radiance contribution of each light source to the scene based on the estimated map, and building the lighting adaptable map by storing the estimated radiance contributions and combining them for any given light setting.

Abstract: A method and a system for calibrating a network of multiple range finders are disclosed. In an embodiment, the method includes performing a measurement with the range finders during a movement of an object through an area covered by the network, identifying, for each range finder, data sets of the measurement associated with the moving object, based on a static analysis of the respective range finder, determining, for each pair of overlapping range finders, an estimated relative transformation between respective poses of the pair, based on the identified data sets of the pair, determining an initial maximum likelihood configuration of poses of all of the range finders based on the estimated relative transformations of each pair and iteratively determining sets of point correspondences of the identified data sets and if a distance of the points in the pair is below a threshold, the threshold being redefined decreasingly for each iteration.

Abstract: A method and a system for calibrating a network of multiple range finders are disclosed. In an embodiment, the method includes performing a measurement with the range finders during a movement of an object through an area covered by the network, identifying, for each range finder, data sets of the measurement associated with the moving object, based on a static analysis of the respective range finder, determining, for each pair of overlapping range finders, an estimated relative transformation between respective poses of the pair, based on the identified data sets of the pair, determining an initial maximum likelihood configuration of poses of all of the range finders based on the estimated relative transformations of each pair and iteratively determining sets of point correspondences of the identified data sets and if a distance of the points in the pair is below a threshold, the threshold being redefined decreasingly for each iteration.

Abstract: To carry out satisfactory object recognition in a short time. An object recognition method in accordance with an exemplary aspect of the present invention is an object recognition method for recognizing a target object by using a preliminarily-created object model. The object recognition method generates a range image of an observed scene, detects interest points from the range image, extracts first features, the first features being features of an area containing the interest points, carries out a matching process between the first features and second features, the second features being features of an area in the range image of the object model, calculates a transformation matrix based on a result of the matching process, the transformation matrix being for projecting the second features on a coordinate system of the observed scene, and recognizes the target object with respect to the object model based on the transformation matrix.

Abstract: To carry out satisfactory object recognition in a short time. An object recognition method in accordance with an exemplary aspect of the present invention is an object recognition method for recognizing a target object by using a preliminarily-created object model. The object recognition method generates a range image of an observed scene, detects interest points from the range image, extracts first features, the first features being features of an area containing the interest points, carries out a matching process between the first features and second features, the second features being features of an area in the range image of the object model, calculates a transformation matrix based on a result of the matching process, the transformation matrix being for projecting the second features on a coordinate system of the observed scene, and recognizes the target object with respect to the object model based on the transformation matrix.