利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

7 条记录 (耗时 0.012 秒)

    TRANSFORMATION OF DATA SAMPLES TO NORMAL DATA
    1.
    发明申请
    TRANSFORMATION OF DATA SAMPLES TO NORMAL DATA 有权

    公开(公告)号：US20220318623A1

    公开(公告)日：2022-10-06

    申请号：US17642212

    申请日：2020-09-22

    申请人： ANOTHER BRAIN

    发明人： David DEHAENE ,  Oriel FRIGO ,  Sébastien COMBREXELLE ,  Pierre ELINE

    IPC分类号： G06N3/08 ,  G06V10/80

    摘要： A device comprising at least one processing logic configured for: obtaining an input vector representing an input data sample; until a stop criterion is met, performing successive iterations of: using an autoencoder trained using a set of reference vectors to encode the input vector into a compressed vector, and decode the compressed vector into a reconstructed vector; calculating a reconstruction loss between the reconstructed and the input vectors, and a gradient of the reconstruction loss; updating said input vector for the subsequent iteration using said gradient.

    Automated method and device capable of providing dynamic perceptive invariance of a space-time event with a view to extracting unified semantic representations therefrom
    4.
    发明授权
    Automated method and device capable of providing dynamic perceptive invariance of a space-time event with a view to extracting unified semantic representations therefrom 有权

    公开(公告)号：US11164049B2

    公开(公告)日：2021-11-02

    申请号：US16499820

    申请日：2018-04-27

    申请人： ANOTHER BRAIN

    发明人： Patrick Pirim

    IPC分类号： G06K9/52 ,  G06F40/30 ,  G06K9/32 ,  G06K9/62 ,  G06N5/02

    摘要： Automated method and device suitable for ensuring the dynamic perceptual invariance of an event with a view to extracting therefrom unified semantic representations are provided. The event is perceived by a linguistic data translator that delivers a signal (HD) referenced (x,y), which signal is transformed into a signal (MAP1) referenced (i,j) through a unit (Dec) that carries out a Gaussian filtering operation that is parameterized by w and decimated by a coefficient k, and transformed into a signal (MAP2) referenced (X,Y) representative of the invariant event through a unit (ROI). A unit (71′) transforms the signal (MAP1) into a signal of oriented edge (ImBo) and of curvature (ImCb), which are presented to a dynamic attractor (80_0) that statistically transforms this information into a curvature cb0 and average orientation bo0, centroid i0,j0 and size of the point cloud ap0,bp0 in order to supply an invariance-computing unit (Inv), which delivers the parameters w, k and the addresses (X,Y) of the signal (MAP2). The invention is applicable in the neuroscience field as an electronically integratable memory-storage unifier.

    ANOMALY DETECTOR, METHOD OF ANOMALY DETECTION AND METHOD OF TRAINING AN ANOMALY DETECTOR
    6.
    发明申请
    ANOMALY DETECTOR, METHOD OF ANOMALY DETECTION AND METHOD OF TRAINING AN ANOMALY DETECTOR 有权

    公开(公告)号：US20230018848A1

    公开(公告)日：2023-01-19

    申请号：US17780490

    申请日：2021-02-03

    申请人： ANOTHER BRAIN

    发明人： Pierre ELINE ,  David DEHAENE

    IPC分类号： G06T7/00 ,  G06N3/04

    摘要： An anomaly detector uses two neural networks, the first, a general purpose classifying convolutional neural network operates as a teacher neural network, while a second neural network in an auto-encoder type configuration. Each of the two neural networks receives the same input stream, and generates respective feature outputs at different levels, corresponding to different resolutions for image data. The respective outputs of the two neural networks are compared at each level, and the resulting difference values consolidated across the difference levels to obtain a final difference value. In a training phase this difference value is used to drive the determination of the weights and biases of the auto-encoder, so as to obtain a auto-encoder trained for a particular input type, under the influence of the teacher neural network. In an operational mode, the difference value is compared to a threshold to determine whether a particular sample is anomalous or not. In certain embodiments, difference values a different levels may be scaled so as to be superimposed at a common resolution, thereby providing an error map indicating the location of anomalous values across the sample.