Abstract: An object detection device includes a processor configured to calculate, for each of a plurality of regions in a detection range of the sensor represented in the newest sensor signal among a plurality of sensor signals in time-series acquired by a sensor, a confidence indicating a degree of certainty that an object to be detected is represented in the region; track a first object which has been detected, to detect, in the newest sensor signal, a passed region through which the first object has passed; control, for each of the plurality of regions in the newest sensor signal, a confidence threshold according to whether or not the region is included in the passed region, and detect a second object in a region, among the plurality of regions, with respect to which the confidence for the second object is equal to or higher than the confidence threshold.

Abstract: An object detection device includes a processor configured to calculate, for each of a plurality of regions in an input image, a confidence indicating a degree of certainty that an object to be detected is represented; determine that the object is represented in a first region for which the confidence is equal to or higher than a first confidence threshold; calculate an occlusion ratio of an occluded region, in which the object is occluded in the first region by another object, to the first region; and determine that the object to be detected is represented in a second region which at least partially overlaps with the first region, when the occlusion ratio is equal to or higher than an occlusion ratio threshold and the confidence for the second region is equal to or higher than a second confidence threshold which is lower than the first confidence threshold.

Abstract: An object detection device includes a processor configured to: detect, by inputting a sensor signal acquired by a sensor to a neural network, an object existing around a vehicle, wherein the neural network includes an input layer, an output layer, and a plurality of layers connected between the input layer and the output layer, and wherein at least one layer of the plurality of layers includes a plurality of sub networks that include the same structure and execute arithmetic processing in parallel to each other on a signal input to the layer; and control, depending on at least either of an amount of available power or available computational resources, the number of sub networks which are used to detect the object among the plurality of sub networks in each of at least one layer.

Abstract: An object detection device comprising a first detecting part detecting types and positions of objects including a road mirror in an image and a presence of a movable object reflected in the road mirror based on image data of a low resolution image and a second detecting part detecting a type of the movable object in the road mirror and a position of the movable object in the road mirror based on high resolution partial image data obtained by extracting the part of the road mirror from the image if there is a movable object reflected in the road mirror.

Abstract: A training image selection system 1, 1? comprises: an imaging device 21, 22 capturing an outside of a host vehicle to generate images, a storage device 32 storing the images generated by the imaging device, a detecting part 34 configured to detect operation of signaling lights of the host vehicle, and an image selecting part 35 configured to select training images from the images generated by the imaging device. If the detecting part detects operation of signaling lights, the image selecting part is configured to select images generated at timings within a predetermined range from when operation of signaling lights is detected, as training images.

Abstract: An object detection device 30 comprises a position region detecting part 31 using a first neural network to detect a position region of an object in the image, a large attribute identification part 32 configured using a second neural network to identify a large attribute of the object, a small attribute identification part 33 using a third neural network to identify a small attribute of the object, and an object judging part 34 judging a result of detection of the object. The object judging part is configured to judge that a result of identification of the small attribute is the result of detection if a confidence of the result of identification of the small attribute is equal to or more than a threshold value, and judge the result of detection based on a result of identification of the large attribute if the confidence is less than the threshold value.

Abstract: An object recognition apparatus is provided with: a first output device configured to output a provisional position of a first object and a tracking priority by using a neural network; and a second output device configured to receive entries of the first image data, the provisional position and the tracking priority, and configured (i) to perform a time-series tracking process and output a position and a speed, for the first object in which the tracking priority satisfies a predetermined condition, and (ii) to output a position without performing the time-series tracking process, for the first object in which the tracking priority does not satisfy the predetermined condition. The neural network is a learned neural network that has learned by using teacher data, wherein the teacher data includes second image data, and correct answer data.