中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

2329 records (took 0.045 seconds)

    Reference signal generation for lock-in amplifier in high sensitivity gas sensing system
    316.
    发明授权
    Reference signal generation for lock-in amplifier in high sensitivity gas sensing system 有权

    公开(公告)号：US09599556B2

    公开(公告)日：2017-03-21

    申请号：US15089987

    申请日：2016-04-04

    Applicant: NEC Laboratories America, Inc.

    Inventor： Junqiang Hu ,  Ting Wang

    IPC: G01N21/00 ,  G01N21/39

    CPC classification number: G01N21/39 ,  G01J3/28 ,  G01J3/4338 ,  G01J2001/4242 ,  G01N2201/06113 ,  G01N2201/12

    Abstract: A gas sensing system includes a signal generator including a wavelength tunable laser, the signal generator providing a first periodic signal and a second periodic signal, wherein the first periodic signal comprises a wavelength scanning signal and the second periodic signal comprises a modulation signal; an optical signal absorption path which is wavelength selective, wherein the generated signal covers at least one of the absorbance band; a signal detector that uses lock-in detection to detect a second harmonic of the second periodic signal after absorption, the signal detector further including a local reference generator, a multiplier, and a low pass filter; a local reference includes a first path (ref1) that outputs sinusoidal signal with frequency equals to that of the second signal in signal generator, and a second path (ref2) that outputs sinusoidal signal of two times ref1 frequency; and a local reference generator having a first phase shifter that is configurable from 0 to 2π and a second phase shifter that shifts 90-degree, wherein the first phase shifter is for an alignment of ref1 with the modulation signal and the second phase shifter provides 90-degree shifts for ref2 from ref1, wherein the first and second paths (ref1 and ref2) are selected by a switch, wherein the switch uses the first path during initialization and the second path for normal operation.

    Annealed Sparsity Via Adaptive and Dynamic Shrinking
    317.
    发明申请
    Annealed Sparsity Via Adaptive and Dynamic Shrinking 审中-公开
    通过自适应和动态收缩退火稀疏

    公开(公告)号：US20160358104A1

    公开(公告)日：2016-12-08

    申请号：US15160280

    申请日：2016-05-20

    Applicant: NEC Laboratories America, Inc.

    Inventor： Kai Zhang ,  Zhengzhang Chen ,  Haifeng Chen ,  Guofei Jiang

    IPC: G06N99/00 ,  G06F17/50

    CPC classification number: G06N20/00 ,  G16B40/20 ,  G16H30/20 ,  G16H40/63 ,  G16H50/20 ,  G16H50/50

    Abstract: Systems and methods are provided for acquiring data from an input signal using multitask regression. The method includes: receiving the input signal, the input signal including data that includes a plurality of features; determining at least two computational tasks to analyze within the input signal; regularizing all of the at least two tasks using shared adaptive weights; performing a multitask regression on the input signal to create a solution path for all of the at least two tasks, wherein the multitask regression includes updating a model coefficient and a regularization weight together under an equality norm constraint until convergence is reached, and updating the model coefficient and regularization weight together under an updated equality norm constraint that has a greater l1-penalty than the previous equality norm constraint until convergence is reached; selecting a sparse model from the solution path; constructing an image using the sparse model; and displaying the image.

    Abstract translation: 提供了系统和方法，用于使用多任务回归从输入信号中获取数据。 所述方法包括：接收所述输入信号，所述输入信号包括包括多个特征的数据; 确定在输入信号内分析的至少两个计算任务; 使用共享自适应权重对所有至少两个任务进行规则化; 对输入信号执行多任务回归，以创建用于所有至少两个任务的解决路径，其中所述多任务回归包括在等式范数约束下一起更新模型系数和正则化权重直到达到收敛，并且更新所述模型 系数和正则化权重在更新的等式规范约束下一起，其具有比先前的等式范数约束更大的l1惩罚，直到达到收敛; 从解决路径中选择稀疏模型; 使用稀疏模型构建图像; 并显示图像。

    ULTRA-HIGH SPEED OPTICAL TRANSPORT EMPLOYING LDPC-CODED MODULATION WITH NON-UNIFORM SIGNALING
    318.
    发明申请
    ULTRA-HIGH SPEED OPTICAL TRANSPORT EMPLOYING LDPC-CODED MODULATION WITH NON-UNIFORM SIGNALING 有权
    超高速光学运输采用非均匀信号的LDPC编码调制

    公开(公告)号：US20160315704A1

    公开(公告)日：2016-10-27

    申请号：US15138184

    申请日：2016-04-25

    Applicant: NEC Laboratories America, Inc.

    Inventor： Ivan DJORDJEVIC ,  Ting WANG

    IPC: H04B10/25 ,  H03M13/11 ,  H04L1/00

    CPC classification number: H04L1/0058 ,  H03M13/1102 ,  H04B10/516 ,  H04L1/0041 ,  H04L1/0057

    Abstract: A low-density parity-check (LDPC) coded bit-interleaved coded modulation with iterative decoding (BICM-ID) scheme with nonuniform signaling which is effected by mapping simple variable-length prefix codes onto the constellation. By employing Huffman procedure(s), prefix codes can be designed to approach optimal performance. Experimental evaluations of the schemes demonstrate that the nonuniform scheme performs better than 8-QAM by at least 8.8 dB.

    Abstract translation: 具有不均匀信令的迭代解码（BICM-ID）方案的低密度奇偶校验（LDPC）编码比特交织编码调制，其通过将简单的可变长度前缀码映射到星座上来实现。 通过采用霍夫曼程序，可以设计前缀码来达到最佳性能。 这些方案的实验评估表明，不均匀方案比8-QAM优于至少8.8dB。

    Non-binary LDPC coded mode-multiplexed four-dimensional signaling based on orthogonal frequency division multiplexing
    319.
    发明授权
    Non-binary LDPC coded mode-multiplexed four-dimensional signaling based on orthogonal frequency division multiplexing 有权
    基于正交频分复用的非二进制LDPC编码模式复用四维信令

    公开(公告)号：US09479285B2

    公开(公告)日：2016-10-25

    申请号：US14513671

    申请日：2014-10-14

    Applicant: NEC Laboratories America, Inc.

    Inventor： Ivan B. Djordjevic ,  Ting Wang

    IPC: H03M13/00 ,  H03M13/05 ,  H04J14/04 ,  H04L1/00 ,  H04L27/26 ,  H04L25/02 ,  H04B7/10 ,  H04L5/00

    CPC classification number: H04J14/04 ,  H04B7/10 ,  H04L1/0041 ,  H04L1/0057 ,  H04L1/0058 ,  H04L5/0026 ,  H04L25/022 ,  H04L25/0224 ,  H04L25/0256 ,  H04L27/2627 ,  H04L27/2634 ,  H04L27/2649

    Abstract: Systems and methods for encoding streams of input data using at least two nonbinary low density parity check (NB-LDPC) encoders; generating NB-LDPC coded optimum signal constellations; performing orthogonal frequency division multiplexing (OFDM) on the NB-LDPC coded four-dimensional (4-D) optimum signal constellations; generating signals using mappers, the mappers configured to assign bits of signals to the signal constellations and to associate the bits of the one or more signals with signal constellation points. Output of the 4-D mappers is modulated using a 4-D OFDM transmitter and a 4-D modulator onto a transmission medium using block coded-modulation, and the modulated output is transmitted by mode-multiplexing independent 4-D OFDM data streams onto fiber. The transmitted modulated output is received, mode-demultiplexed, and demodulated using polarization diversity receivers, one per spatial mode, channel estimation and compensation methods are performed to overcome impairments in the transmission medium; and received data is decoded using non-binary decoders.

    Abstract translation: 使用至少两个非二进制低密度奇偶校验（NB-LDPC）编码器对输入数据流进行编码的系统和方法; 生成NB-LDPC编码的最佳信号星座; 对NB-LDPC编码的四维（4-D）最佳信号星座进行正交频分复用（OFDM）; 使用映射器产生信号，映射器被配置为将信号位分配给信号星座并将一个或多个信号的位与信号星座点相关联。 使用块编码调制，使用4-D OFDM发送器和4-D调制器将4-D映射器的输出调制到传输介质上，并且经调制的输出通过将独立的4-D OFDM数据流模式复用传输到 纤维。 传输的调制输出被接收，模式解复用和使用极化分集接收机解调，每个空间模式一个，信道估计和补偿方法被执行以克服传输介质中的损伤; 并且使用非二进制解码器对接收的数据进行解码。

    Fine-grained Image Classification by Exploring Bipartite-Graph Labels
    320.
    发明申请
    Fine-grained Image Classification by Exploring Bipartite-Graph Labels 审中-公开
    通过探索双边图标签进行细粒度图像分类

    公开(公告)号：US20160307072A1

    公开(公告)日：2016-10-20

    申请号：US15095260

    申请日：2016-04-11

    Applicant: NEC Laboratories America, Inc.

    Inventor： Feng Zhou ,  Yuanqing Lin

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

    CPC classification number: G06K9/66 ,  G06K9/469 ,  G06K9/6267 ,  G06K9/627 ,  G06N3/0454 ,  G06N3/0472 ,  G06N3/084 ,  G06T7/0004 ,  G06T2207/30128

    Abstract: Systems and methods are disclosed for deep learning and classifying images of objects by receiving images of objects for training or classification of the objects; producing fine-grained labels of the objects; providing object images to a multi-class convolutional neural network (CNN) having a softmax layer and a final fully connected layer to explicitly model bipartite-graph labels (BGLs); and optimizing the CNN with global back-propagation.

    Abstract translation: 公开了用于深入学习和分类对象的图像的系统和方法，通过接收用于对象的训练或分类的对象的图像; 生产物品的细粒标签; 向具有softmax层和最终完全连接的层的多级卷积神经网络（CNN）提供对象图像以明确地模拟二分图标签（BGL）; 并利用全局反向传播优化CNN。

Patent Agency Ranking