Fujitsu Corporation

Nannan Wang

Paparao Palacharla

Xi Wang

Tadashi Ikeuchi

Qiong Zhang

Weisheng Xie

Jian Kong

Qi Chen

Soroush Bateni

A method may include obtaining an image of a scene from a first perspective, the image including an object, and detecting the object in the image using a machine learning process, where the object may be representative of a known shape with at least four vertices at a first set of points. The method may also include automatically predicting a second set of points corresponding to the at least four vertices of the object in a second perspective of the scene based on the known shape of the object. The method may additionally include constructing, without user input, a transformation matrix to transform a given image from the first perspective to the second perspective based on the first set of points and the second set of points.

Automatic perspective transformation

Systems and methods including identifying a physical intelligent moving object (IMO), the physical IMO associated with a current location and a service request; creating, from a digital representation template that is based on the service request, a digital representation instance of the physical IMO in a particular edge cloud of a distributed cloud computing environment that is closest to the current location of the physical IMO; in response to creating the digital representation instance, establishing a network connection between the digital representation and the physical IMO using the particular edge cloud; detecting a movement of the physical IMO from coverage of the particular edge cloud; in response to detecting the movement of the physical IMO: identifying a target edge cloud of the distributed cloud computing environment for coverage of the physical IMO; and establishing a network connection between the digital representation and the physical IMO using the target edge cloud.

Digital representations of physical intelligent moving objects

Systems and methods for identifying physical intelligent moving objects (IMOs); creating, for a first subset of the physical IMOs, digital representations in edge clouds of a distributed cloud computing environment, each digital representation representative of a physical IMO of the first subset of the physical IMOs; generating a cooperative network computing (CNC) platform that includes the digital representations of the first subset of the physical IMOs, each digital representation functioning as a CNC entity in the CNC platform for the corresponding physical IMO; identifying a CNC task of a first physical IMO of the plurality of physical IMOs; and performing, by a first digital representation of the representations that corresponds to the first physical IMO, the CNC task in the CNC platform, including accessing data associated with a second digital representation of the digital representations in the CNC platform to perform the CNC task at the first digital representation.

Edge-accelerated artificial intelligence for collaborative sensing of physical intelligent moving objects

Disclosed methods for network modernization include obtaining a list of end-to-end circuits carried in a circuit-switched network, calculating, for each circuit, an early retirement credit (ERC) score and a circuit load factor (CLF) score, selecting, dependent on the ERC and CLF scores, a circuit to migrate to a new network, adding the selected circuit to a circuit migration sequence, and removing the circuit from the list. The ERC score represents the number of circuit-switching units on which no circuits would be carried and that would remain in the network following its removal. The CLF score represents an average number of circuits that would be carried on each circuit-switching unit currently traversed by the circuit following its removal. When two circuits have the highest ERC score, the circuit with the lowest CLF score is selected for migration. The method is repeated until the list is empty.

Circuit migration sequence generation for network modernization

Systems and methods for provisioning real-time three-dimensional maps, including: at a first processing stage: updating a three-dimensional map of an environment based on received point cloud data; extracting three-dimensional proposals of objects of the environment; projecting the three-dimensional proposals onto a two-dimensional image of the environment; generating two-dimensional proposals of the objects of the environment; at a second processing stage: detecting the objects of the environment from the two-dimensional image of the environment; generating two-dimensional bounding boxes of the objects of the environment; matching the two-dimensional bounding box with a particular two-dimensional proposal of the two-dimensional proposals; and labeling, for each two-dimensional proposal that is matched to a two-dimensional bounding box, the three-dimensional proposal corresponding to the two-dimensional proposal with semantic information associated with the two-dimensional bounding box that is matched to the two-dimensional proposal to update the three-dimensional map.

Provisioning real-time three-dimensional maps for autonomous vehicles

Systems and methods for transmitting, within a time slot, i) first eMBB resource blocks to a first UE and ii) second eMBB resource blocks to a second UE; during transmission of the first and the second eMBB resource blocks to the first and the second UE, respectively, at a first scheduled check point: identifying i) first URLLC resource blocks to be transmitted to a third UE and ii) second URLLC resource blocks to be transmitted to a fourth UE; identifying i) a first transmission deadline for transmitting the first URLLC resource blocks to the third UE and ii) a second transmission deadline for transmitting the second URLLC resource blocks to the fourth UE; generating a transmission queue for the first URLLC resource blocks and the second URLLC resource blocks; and selecting, based on the transmission queue, the second URLLC resource blocks for transmission to the fourth UE.

Multiplexing of resource blocks within a communications network for dynamic service provisioning

An intersection management system (IMS) may receive one or more traversing requests from one or more Connected Autonomous Vehicles (CAVs). The IMS may determine a solution space for each of the one or more traversing requests in a space-time resource model of the intersection, find a CAV trajectory allocation in the space-time resource model for each of the one or more traversing requests. The IMS may send an approved reservation to each CAV corresponding to each of the one or more CAV trajectory allocations that have been found. Each of the one or more CAVs may, when an approved reservation corresponding to the CAV may have been received from the IMS, move through the intersection zone as specified in the approved reservation.

Fragmentation-aware intelligent autonomous intersection management using a space-time resource model

Systems and methods for identifying a service region that includes sub-regions; calculating, for each pair of neighboring sub-regions, traffic data of computing devices traveling between the pair of sub-region; generating a directed graph representing the service region, the directed graph including nodes representing the sub-regions and the links between the nodes representing the first and the second traffic data; partitioning the directed graph into graph partitions, each graph partition including sub-graphs; calculating, for each sub-graph of each graph partition, a net traffic flow for the sub-graph based on a difference of the first and the second traffic data of the sub-regions that correspond to the nodes of the sub-graph; calculating, for each graph partition, a net traffic flow for the graph partition based on a summation of the net traffic flow of each sub-graph of the graph partition; identifying a particular graph partition having a smallest net traffic flow.

Distribution of computing resources of a radio access network

Systems and methods for traffic congestion avoidance may use cooperative autonomous driving based on vehicle-to-vehicle communication to resolve deadlock conditions. The method may include automatically detecting, by a given autonomous vehicle, that the given vehicle is deadlocked, initiating, by the given vehicle, formation of a coordination group including the given vehicle, a front vehicle traveling in the same driving lane as the given vehicle, and an assistant autonomous vehicle traveling in a lane adjacent to the lane in which the given vehicle and the front vehicle are traveling, and sending, by the given vehicle using vehicle-to-vehicle communication, a cooperative driving task request to initiate actions by the autonomous vehicles in the coordination group to resolve the deadlock condition. The actions may include modifying a driving plan of an autonomous vehicle in the coordination group and initiating a speed negotiation with an autonomous vehicle in the coordination group.

Growing community of inventors

Allen, TX, United States of America

Nannan Wang