公开(公告)号：US20180307967A1

公开(公告)日：2018-10-25

申请号：US15785796

申请日：2017-10-17

Applicant: NEC Laboratories America, Inc.

Inventor： Hans Peter Graf ,  Eric Cosatto ,  Iain Melvin

IPC: G06N3/04 ,  G06N3/08 ,  G01S7/41

CPC classification number: G06N3/04 ,  G01S7/417 ,  G01S13/931 ,  G06N3/08

Abstract: A computer-implemented method executed by a processor for training a neural network to recognize driving scenes from sensor data received from vehicle radar is presented. The computer-implemented method includes extracting substructures from the sensor data received from the vehicle radar to define a graph having a plurality of nodes and a plurality of edges, constructing a neural network for each extracted substructure, combining the outputs of each of the constructed neural networks for each of the plurality of edges into a single vector describing a driving scene of a vehicle, and classifying the single vector into a set of one or more dangerous situations involving the vehicle.

公开(公告)号：US10107651B2

公开(公告)日：2018-10-23

申请号：US15811061

申请日：2017-11-13

Applicant: NEC Laboratories America, Inc.

Inventor： Fatih Yaman ,  Shaoliang Zhang ,  Ezra Ip ,  Yue-Kai Huang ,  Giovanni Milione

IPC: G01N21/00 ,  G01D5/353 ,  G01M11/00

Abstract: Methods and systems for Brillouin optical time-domain reflectometry include optically filtering out non-Brillouin signals reflected from a fiber. Brillouin signals reflected from the fiber are coupled with a local oscillator to produce a Brillouin shift signals. The Brillouin shift signals are converted to an electrical domain using a photodetector. The electrical Brillouin shift signals are converted to a digital domain using a low-speed analog-to-digital converter that has a sampling rate below a Nyquist rate sufficient to fully resolve the electrical Brillouin shift signals.

公开(公告)号：US20180299877A1

公开(公告)日：2018-10-18

申请号：US15880979

申请日：2018-01-26

Applicant: NEC Laboratories America, Inc.

Inventor： Wei Cheng ,  Haifeng Chen ,  Kenji Yoshihira

IPC: G05B23/02 ,  G01D5/00 ,  G06F11/07

Abstract: A method, computer program product, and a system is provided for power plant system fault diagnosis. The method includes detecting, using an invariant model, a fault event based on a broken pair-wise correlation. The method also includes constructing a fault signature based on the fault event. The method further includes generating a feature vector in a feature subspace for the fault signature, wherein said feature vector includes at least one status of at least one system component during the fault event. The method additionally includes determining a corrective action correlated to the fault signature, from among a plurality of candidate corrective actions associated with the one or more historical representative signature, based on a Jaccard similarity using the feature vector in the feature subspace. The method also includes initiating the corrective action on a hardware device to mitigate expected harm.

公开(公告)号：US20180270849A1

公开(公告)日：2018-09-20

申请号：US15915491

申请日：2018-03-08

Applicant: NEC Laboratories America, Inc.

Inventor： Karthikeyan Sundaresan ,  Eugene Chai ,  Mohammad Khojastepour ,  Sampath Rangarajan

IPC: H04W72/12 ,  H04W56/00 ,  H04W48/16

Abstract: A system is provided for speculative scheduling that includes a base station having a processor. The processor computes an overall schedule for a set of clients. The overall schedule is formed from a set of speculative schedules, is configured to maximize unlicensed spectrum usage, and is computed by (a) determining a speculative schedule for each resource block from a set of resource blocks in a given sub-frame based on statistics determined for the clients individually and jointly, and (b) selecting, for formation into the overall schedule, (i) a particular resource block and (ii) the speculative schedule for the particular resource block, that yield the maximum incremental utility relative to already determined speculative schedules for other resource blocks in the set, based on criteria including uplink access statistics. The processor executes the overall schedule responsive to a completion of the speculative schedule determination for each resource block in the given sub-frame.

公开(公告)号：US20180268292A1

公开(公告)日：2018-09-20

申请号：US15908870

申请日：2018-03-01

Applicant: NEC Laboratories America, Inc.

Inventor： Wongun Choi ,  Manmohan Chandraker ,  Guobin Chen ,  Xiang Yu

IPC: G06N3/08 ,  G06K9/66

CPC classification number: G06N3/08 ,  G06K9/00684 ,  G06K9/4628 ,  G06K9/6217 ,  G06K9/6264 ,  G06K9/6274 ,  G06K9/66 ,  G06N3/0454 ,  G06N3/0481 ,  G06N3/084

Abstract: A computer-implemented method executed by at least one processor for training fast models for real-time object detection with knowledge transfer is presented. The method includes employing a Faster Region-based Convolutional Neural Network (R-CNN) as an objection detection framework for performing the real-time object detection, inputting a plurality of images into the Faster R-CNN, and training the Faster R-CNN by learning a student model from a teacher model by employing a weighted cross-entropy loss layer for classification accounting for an imbalance between background classes and object classes, employing a boundary loss layer to enable transfer of knowledge of bounding box regression from the teacher model to the student model, and employing a confidence-weighted binary activation loss layer to train intermediate layers of the student model to achieve similar distribution of neurons as achieved by the teacher model.

公开(公告)号：US20180268265A1

公开(公告)日：2018-09-20

申请号：US15889846

申请日：2018-02-06

Applicant: NEC Laboratories America, Inc.

Inventor： Kihyuk Sohn ,  Xiang Yu ,  Manmohan Chandraker

IPC: G06K9/66 ,  G06T7/70 ,  G06T9/00 ,  G06K9/00 ,  G06K9/62

Abstract: An object recognition system is provided that includes a device configured to capture a video sequence formed from unlabeled testing video frames. The system includes a processor configured to pre-train a recognition engine formed from a reference set of CNNs on a still image domain that includes labeled training still image frames. The processor adapts the recognition engine to a video domain to form an adapted recognition engine, by applying a non-reference set of CNNs to a set of domains that include the still image and video domains and a degraded image domain. The degraded image domain includes labeled synthetically degraded versions of the labeled training still image frames included in the still image domain. The video domain includes random unlabeled training video frames. The processor recognizes, using the adapted engine, a set of objects in the video sequence. A display device displays the set of recognized objects.

公开(公告)号：US20180262293A1

公开(公告)日：2018-09-13

申请号：US15912781

申请日：2018-03-06

Applicant: NEC LABORATORIES AMERICA, INC.

Inventor： Ezra Ip ,  Yue-Kai Huang ,  Shuji Murakami ,  Philip Ji

IPC: H04J14/02

CPC classification number: H04J14/0215 ,  H04J14/0204 ,  H04J14/0205 ,  H04J14/0212 ,  H04J14/023 ,  H04J14/0282 ,  H04J14/0295 ,  H04Q11/0005 ,  H04Q2011/0016

Abstract: Aspects of the present disclosure describe systems, methods, and structures for passive optical add/drop multiplexing (POADM) architectures that remove the prior art requirement of an optical amplifier (i.e., repeater-less) at the POADM nodes.

公开(公告)号：US20180261968A1

公开(公告)日：2018-09-13

申请号：US15912810

申请日：2018-03-06

Applicant: NEC LABORATORIES AMERICA, INC.

Inventor： Fatih Yaman ,  Shaoliang Zhang ,  Eduardo Mateo Rodriquez ,  Takanori Inoue ,  Yoshihisa Inada

IPC: H01S3/067 ,  H01S3/10 ,  H01S3/16 ,  H04J14/02

CPC classification number: H01S3/0677 ,  H01S3/06766 ,  H01S3/06787 ,  H01S3/10007 ,  H01S3/1608 ,  H01S3/2383 ,  H01S2301/02 ,  H04B10/294 ,  H04B10/2972 ,  H04J14/02

Abstract: Aspects of the present disclosure describe systems, methods, and structures for providing C-band and L-band transmission exhibiting increased power efficiency by diverting a portion of C-band optical energy to an input of L-band optical amplifiers (C-seeding) while optionally employing circulators to eliminate the need for optical isolators.

公开(公告)号：US20180255385A1

公开(公告)日：2018-09-06

申请号：US15907728

申请日：2018-02-28

Applicant: NEC Laboratories America, Inc.

Inventor： Ivan B. Djordjevic ,  Shaoliang Zhang ,  Ting Wang

IPC: H04Q11/00

CPC classification number: G01D5/35383 ,  G01D5/268 ,  G02B5/32 ,  G02B6/3512 ,  G02B6/3546 ,  G02B6/356 ,  G02B6/3584 ,  G02B26/0833 ,  H04J14/06 ,  H04L63/0428 ,  H04L67/12 ,  H04Q11/0005 ,  H04Q11/0062 ,  H04Q2011/0024 ,  H04Q2011/003 ,  H04Q2011/0064 ,  H04Q2011/0081 ,  H04Q2011/0088 ,  H04Q2011/0094 ,  H04W4/90 ,  H04W84/18

Abstract: A method and network are provided for a reconfigurable optical sensor network. The method includes configuring, by a controller, the reconfigurable optical sensor network, including one or more reconfigurable optical space switches, for a type of sensor data. The method also includes generating sensor data in the type of sensor data with one or more of a plurality of bidirectional sensors. The method additionally includes sending the sensor data to one or more optical star couplers. The method further includes forwarding the sensor data from one of the one or more optical star couplers to the one of one or more reconfigurable optical space switches.

公开(公告)号：US20180213415A1

公开(公告)日：2018-07-26

申请号：US15874016

申请日：2018-01-18

Applicant: NEC Laboratories America, Inc.

Inventor： Mohammad Khojastepour ,  Narayan Prasad ,  Sampath Rangarajan ,  Yasaman Ghasempour

IPC: H04W24/02 ,  H04W16/28 ,  H04W72/04

CPC classification number: H04W24/02 ,  H04B7/0695 ,  H04W16/28 ,  H04W72/044

Abstract: A computer-implemented method executed in a wireless communication system for establishing communication links in a millimeter wave network is presented. The method includes determining active communication links between a plurality of transmitters and a plurality of receivers operated by a plurality of users and formulating a quality of service (QoS) proportional fairness (PF) network utility as a constrained submodular set function maximization problem. The method further includes performing an optimization framework by optimizing, based on the active communication links, user association for fixed transmit beam patterns and optimizing, based on the active communication links, the fixed transmit beam patterns and corresponding beam attributes for a fixed set of user associations. The method also includes iteratively performing the optimizing steps in an alternating manner to determine a user from the plurality of users that optimizes the constrained submodular set function maximization problem and assigning the determined user to a cell.