会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 5. 发明授权
    • Chip with integrated circuit and micro-silicon condenser microphone integrated on single substrate and method for making the same
    • 芯片集成电路和微硅电容麦克风集成在单个基板上,并制作相同的方法
    • US09221675B2
    • 2015-12-29
    • US13561194
    • 2012-07-30
    • Wei HuGang LiJia-Xin Mei
    • Wei HuGang LiJia-Xin Mei
    • B81C1/00H04R31/00H04R19/00
    • B81C1/00158B81C1/00246H04R19/005H04R31/00
    • A method for integrating an IC and a MEMS component includes the following steps: S1) providing a SOI base (20) having a first area (21) and a second area (22); S2) fabricating an IC on the first area through a standard semiconductor process, and simultaneously forming a metal conductive layer (26) and a medium insulation layer (25c) extending to the second area; S3) partly removing the medium insulation layer and then further partly removing the silicon component layer so as to form a backplate diagram; S4) depositing a sacrificial layer (32) above the SOI base; S5) forming a Poly Sil-xGex film (33) on the sacrificial layer; S6) forming a back cavity (34); and S7) eroding the sacrificial layer to form a chamber (36) in communication with the back cavity. Besides, a chip (10) fabricated by the above method is also disclosed.
    • 一种用于集成IC和MEMS部件的方法包括以下步骤:S1)提供具有第一区域(21)和第二区域(22)的SOI基底(20); S2)通过标准半导体工艺在第一区域上制造IC,同时形成延伸到第二区域的金属导电层(26)和介质绝缘层(25c); S3)部分地去除介质绝缘层,然后进一步部分去除硅组分层以形成背板图; S4)在SOI基底上方沉积牺牲层(32); S5)在牺牲层上形成Poly Sil-xGex膜(33); S6)形成后腔(34); 和S7)侵蚀牺牲层以形成与后腔连通的腔室(36)。 此外,还公开了通过上述方法制造的芯片(10)。
    • 7. 发明授权
    • Controllers for DC to DC converters
    • DC到DC转换器的控制器
    • US09059632B2
    • 2015-06-16
    • US12874438
    • 2010-09-02
    • Gang LiFengjiang ZhangLaszlo Lipcsei
    • Gang LiFengjiang ZhangLaszlo Lipcsei
    • G05F1/00H02M3/158H02M1/00
    • H02M3/1588H02M2001/0009H02M2001/0032Y02B70/1466Y02B70/16
    • A controller includes a ramp signal generator and control circuitry coupled to the ramp signal generator. The ramp signal generator provides a control current through a resistive component to control energy stored in a first energy storage component. The ramp signal generator further generates a ramp signal based on the energy stored in the first energy storage component. The control circuitry adjusts a voltage at one end of the resistive component thereby controlling the control current to indicate a voltage across a second energy storage component. The control circuitry further controls a current through the second energy storage component within a predetermined range based on the ramp signal.
    • 控制器包括斜坡信号发生器和耦合到斜坡信号发生器的控制电路。 斜坡信号发生器通过电阻部件提供控制电流,以控制存储在第一能量存储部件中的能量。 斜坡信号发生器还基于存储在第一能量存储部件中的能量产生斜坡信号。 控制电路调节电阻部件一端的电压,从而控制控制电流以指示跨越第二能量存储部件的电压。 控制电路还基于斜坡信号进一步控制在预定范围内通过第二能量存储部件的电流。
    • 9. 发明授权
    • Exploiting sparseness in training deep neural networks
    • 在深层神经网络训练中利用稀疏性
    • US08700552B2
    • 2014-04-15
    • US13305741
    • 2011-11-28
    • Dong YuLi DengFrank Torsten Bernd SeideGang Li
    • Dong YuLi DengFrank Torsten Bernd SeideGang Li
    • G06F15/18G06N3/08
    • G06N3/08
    • Deep Neural Network (DNN) training technique embodiments are presented that train a DNN while exploiting the sparseness of non-zero hidden layer interconnection weight values. Generally, a fully connected DNN is initially trained by sweeping through a full training set a number of times. Then, for the most part, only the interconnections whose weight magnitudes exceed a minimum weight threshold are considered in further training. This minimum weight threshold can be established as a value that results in only a prescribed maximum number of interconnections being considered when setting interconnection weight values via an error back-propagation procedure during the training. It is noted that the continued DNN training tends to converge much faster than the initial training.
    • 提出了深层神经网络(DNN)训练技术实施例,其训练DNN,同时利用非零隐藏层互连权重值的稀疏性。 通常,完全连接的DNN最初通过遍历完整的训练集多次进行训练。 那么,在大多数情况下,只有重量大小超过最小重量阈值的互连在进一步的训练中被考虑。 该最小权重阈值可以被建立为在训练期间通过错误反向传播过程设置互连权重值时仅考虑规定的最大数量的互连的值。 值得注意的是,继续进行的DNN训练往往比初始训练快得多。
    • 10. 发明授权
    • Prodrugs based on gemcitabine structure and synthetic methods and applications thereof
    • 基于吉西他滨结构的前体药物及其合成方法及应用
    • US08653048B2
    • 2014-02-18
    • US13265335
    • 2010-03-25
    • Xiaoxia XueGang LiChangjun SunWenbao Li
    • Xiaoxia XueGang LiChangjun SunWenbao Li
    • A61K31/70C07H19/00
    • C07H19/06A61K31/7068
    • Prodrugs based on gemcitabine structure shown in formula (I) as well as their synthetic method and application are disclosed in the present invention, wherein the definitions for the groups of a, b, c, d, E, Z and V are described in the specification. By modifying the N4 group, the solubility, the bioavailability and the organ specificity of the prodrugs are improved. Therefore, the fast metabolism problem is overcome for the produced prodrugs compounds. Intestinal toxicity induced by gemcitabine is decreased. Thereby, the prodrugs can be delivered by oral administration in clinics and further improve their anti-tumor, anti-cancer, anti-infection and diffusion preventing capability, and can also specifically act on liver or colon. The synthetic method is simple and adapted to industrial production.
    • 基于式(I)所示的吉西他滨结构的前药以及它们的合成方法和应用在本发明中公开,其中a,b,c,d,E,Z和V组的定义描述于 规范。 通过改变N4基团,前药的溶解度,生物利用度和器官特异性得到改善。 因此,所制备的前药化合物克服了快速代谢问题。 吉西他滨引起的肠毒性降低。 因此,前药可以通过口服给药在临床中进一步提高其抗肿瘤,抗癌,抗感染和扩散防止能力,并且还可以特异性地作用于肝脏或结肠。 合成方法简单,适应工业生产。