Abstract: A learning apparatus performs a learning process for a feed-forward multilayer neural network with supervised learning. The network includes an input layer, an output layer, and at least one hidden layer having at least one probing neuron that does not transfer an output to an uppermost layer side of the network. The learning apparatus includes a learning unit and a layer quantity adjusting unit. The learning unit performs a learning process by calculation of a cost derived by a cost function defined in the multilayer neural network using a training data set for supervised learning. The layer quantity adjusting unit removes at least one uppermost layer from the network based on the cost derived by the output from the probing neuron, and sets, as the output layer, the probing neuron in the uppermost layer of the remaining layers.

Abstract: The moving object recognition system includes: a camera that is installed in a vehicle and captures continuous single-view images; a moving object detecting unit that detects a moving object from the images captured by the camera; a relative approach angle estimating unit that estimates the relative approach angle of the moving object detected by the moving object detecting unit with respect to the camera; a collision risk calculating unit that calculates the risk of the moving object colliding with the vehicle, based on the relationship between the relative approach angle and the moving object direction from the camera toward the moving object; and a reporting unit that reports a danger to the driver of the vehicle in accordance with the risk calculated by the collision risk calculating unit.

Abstract: A detection device has a neural network process section performing a neural network process using parameters to calculate and output a classification result and a regression result of each of frames in an input image. The classification result shows a presence of a person in the input image. The regression result shows a position of the person in the input image. The parameters are determined based on a learning process using a plurality of positive samples and negative samples. The positive samples have segments of a sample image containing at least a part of the person and a true value of the position of the person in the sample image. The negative samples have segments of the sample image containing no person.

Abstract: When a driving support apparatus is under automatic driving of a vehicle or an automatic driving button is pressed under manual driving, vicinity image data is acquired from an in-vehicle camera. When a predetermined target object is recognized in the vicinity image data, a visibility reduction process is applied to image data of the recognized target object. The visibility reduction process applies at least one of defocusing; decreasing color information; and decreasing edge intensity, to the image data of the recognized target object. In contrast, any visibility reduction process is not applied to any other image data other than the image data of the recognized target object. An image display apparatus displays the vicinity image by a combination of the image data of the recognized target object of which the visibility is reduced and the other image data of which the visibility is not reduced.

Abstract: A detection device has a neural network process section performing a neural network process using parameters to calculate and output a classification result and a regression result of each of frames in an input image. The classification result shows a presence of a person in the input image. The regression result shows a position of the person in the input image. The parameters are determined based on a learning process using a plurality of positive samples and negative samples. The positive samples have segments of a sample image containing at least a part of the person and a true value of the position of the person in the sample image. The negative samples have segments of the sample image containing no person.

Abstract: A pedestrian notification apparatus, which is configured to be used in a vehicle, includes an ambient object detector, a notification target detector and a notifier. The ambient object detector is configured to detect ambient objects present around the vehicle. The notification target detector is configured to detect a notification target pedestrian based on the detection results of the ambient object detector. The notifier is configured to notify the notification target pedestrian that the notification target pedestrian has been detected by the notification target detector.

Abstract: A learning apparatus performs a learning process for a feed-forward multilayer neural network with supervised learning. The network includes an input layer, an output layer, and at least one hidden layer having at least one probing neuron that does not transfer an output to an uppermost layer side of the network. The learning apparatus includes a learning unit and a layer quantity adjusting unit. The learning unit performs a learning process by calculation of a cost derived by a cost function defined in the multilayer neural network using a training data set for supervised learning. The layer quantity adjusting unit removes at least one uppermost layer from the network based on the cost derived by the output from the probing neuron, and sets, as the output layer, the probing neuron in the uppermost layer of the remaining layers.

Abstract: A discriminator based on supervised learning includes a data expanding unit and a discriminating unit. The data expanding unit performs data expansion on unknown data which is an object to be discriminated in such a manner that a plurality of pieces of pseudo known data are generated. The discriminating unit applies the plurality of pieces of unknown pseudo data that has been expanded by the data expansion unit to a discriminative model so as to discriminate the plurality of pieces of pseudo unknown data, and integrates discriminative results of the plurality of pieces of pseudo unknown data to perform class classification such that the unknown data is classified into classes.

Abstract: When a driving support apparatus is under automatic driving of a vehicle or an automatic driving button is pressed under manual driving, vicinity image data is acquired from an in-vehicle camera. When a predetermined target object is recognized in the vicinity image data, a visibility reduction process is applied to image data of the recognized target object. The visibility reduction process applies at least one of defocusing; decreasing color information; and decreasing edge intensity, to the image data of the recognized target object. In contrast, any visibility reduction process is not applied to any other image data other than the image data of the recognized target object. An image display apparatus displays the vicinity image by a combination of the image data of the recognized target object of which the visibility is reduced and the other image data of which the visibility is not reduced.

Abstract: When it is determined that a minimum value of a cost function of a candidate structure obtained by a training process of a specified-number sequence is equal to or higher than that of the cost function of the candidate structure obtained by the first step of a previous sequence immediately before the specified-number sequence, a method performs, as a random removal step of the specified sequence, a step of randomly removing at least one unit from the candidate structure obtained by the first step of the previous sequence again. This gives a new generated structure of the target neural network based on the random removal to the first step as the input structure of the target neural network. The method performs the specified-number sequence again using the new generated structure of the target neural network.

Abstract: A symbolizing execution section symbolizes information regarding driving scenes, and describes an entire driving scene around the location of an own motor vehicle. This makes it possible to only increase the information to be symbolized when the number of traffic participants is increased. This can adopt the increase of the traffic participants. A symbol transition prediction section predicts a transition of the symbolized driving scene by using influence which affects the operation of the traffic participants caused by the operation of the own motor vehicle. This makes it possible to increase prediction accuracy while decreasing a large amount of calculation on predicting the transition of driving scene.

Abstract: A pedestrian notification apparatus, which is configured to be used in a vehicle, includes an ambient object detector, a notification target detector and a notifier. The ambient object detector is configured to detect ambient objects present around the vehicle. The notification target detector is configured to detect a notification target pedestrian based on the detection results of the ambient object detector. The notifier is configured to notify the notification target pedestrian that the notification target pedestrian has been detected by the notification target detector.

Abstract: The moving object recognition system includes: a camera that is installed in a vehicle and captures continuous single-view images; a moving object detecting unit that detects a moving object from the images captured by the camera; a relative approach angle estimating unit that estimates the relative approach angle of the moving object detected by the moving object detecting unit with respect to the camera; a collision risk calculating unit that calculates the risk of the moving object colliding with the vehicle, based on the relationship between the relative approach angle and the moving object direction from the camera toward the moving object; and a reporting unit that reports a danger to the driver of the vehicle in accordance with the risk calculated by the collision risk calculating unit.

Abstract: A symbolizing execution section symbolizes information regarding driving scenes, and describes an entire driving scene around the location of an own motor vehicle. This makes it possible to only increase the information to be symbolized when the number of traffic participants is increased. This can adopt the increase of the traffic participants. A symbol transition prediction section predicts a transition of the symbolized driving scene by using influence which affects the operation of the traffic participants caused by the operation of the own motor vehicle. This makes it possible to increase prediction accuracy while decreasing a large amount of calculation on predicting the transition of driving scene.

Abstract: A camera of a ground-based information dispatch apparatus captures a blind-spot image, showing a region that is a blind spot with respect to a vehicle driver. A vehicle-mounted camera captures a forward-view image corresponding to the viewpoint of the driver, and the forward-view image is transmitted to the information dispatch apparatus together with vehicle position and direction information and camera parameters. Based on the received information, the blind-spot image is converted to a corresponding image having the viewpoint of the vehicle driver, and the forward-view image and viewpoint-converted blind-spot image are combined to form a synthesized image, which is transmitted to the vehicle.

Abstract: A road surface reflection detecting apparatus includes a road recognition unit for recognizing a road in a picture taken by a camera of a road existing in front of a vehicle. A road surface reflection detection unit extracts a picture from the road recognized by the road recognition unit and determines the degree of reflection of light from the surface of the road based on the picture extracted from the road. The picture includes the reflection of light beam radiated from vehicle headlights. If the reflection is determined to be high, the direction or intensity of the light beam from the vehicle headlights is regulated.

Abstract: An image processor for an automotive vehicle includes: an imaging element for capturing an image of a road; a traffic lane line detection element for detecting a position of a traffic lane line by determining an edge of the traffic lane line; a distance distribution calculation element for calculating a distance distribution of an edge intensity corresponding to a distance to the traffic lane line from the imaging element; and a detection reliability calculation element for calculating a degree of reliability of detection of the traffic lane line on the basis of the distance distribution of the edge intensity of the traffic lane line.

Abstract: A camera of a ground-based information dispatch apparatus captures a blind-spot image, showing a region that is a blind spot with respect to a vehicle driver. A vehicle-mounted camera captures a forward-view image corresponding to the viewpoint of the driver, and the forward-view image is transmitted to the information dispatch apparatus together with vehicle position and direction information and camera parameters. Based on the received information, the blind-spot image is converted to a corresponding image having the viewpoint of the vehicle driver, and the forward-view image and viewpoint-converted blind-spot image are combined to form a synthesized image, which is transmitted to the vehicle.

Abstract: In a vehicle headlight apparatus, a projection unit having a same structure of an LED projector receives a light provided from a lamp. The projection unit generates and projects a projection image toward a vehicle front-area. A vehicle ECU has an image signal obtaining means for obtaining image signals representing a vehicle front-area image, a high brightness area determination means for analyzing the image signals in order to successively determine a high brightness area in the image signals, and a projection image determination means for determining a projection image so that the light, corresponding to the high brightness area and to be irradiated toward the vehicle front-area, becomes weak when compared with the light previously determined. Thereby, an intensity of the light to be irradiated to an object with a high reflection factor can be decreased. This can suppress decreasing of driver's frontal visibility.

Abstract: A lane detection apparatus includes a capture means, a vehicle detection means, a subject lane detection means, and an additional lane detection means. The capture means captures an image around a subject vehicle. The vehicle detection means detects an object vehicle around the subject vehicle. The subject lane detection means detects a subject lane. The additional lane detection means detects an additional lane different from the subject lane based on a positional relationship between the object vehicle detected by the vehicle detection means and the subject lane detected by the subject lane detection means.