利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

23 条记录 (耗时 0.016 秒)

    Parameter redundancy reduction method
    3.
    发明授权
    Parameter redundancy reduction method 有权

    公开(公告)号：US11972108B2

    公开(公告)日：2024-04-30

    申请号：US17525999

    申请日：2021-11-15

    申请人： INTERNATIONAL BUSINESS MACHINES CORPORATION

    发明人： Zhong Fang Yuan ,  Tong Liu ,  Li Juan Gao ,  Na Liu ,  Xiang Yu Yang

    IPC分类号： G06F3/06 ,  G06N3/04

    CPC分类号： G06F3/0608 ,  G06F3/0641 ,  G06F3/0673 ,  G06N3/04

    摘要： A method, computer program product, and computer system for generating and using a basic state layer. N task models are provided (N≥2). Each task model was trained on a same pre-trained backbone model. Each task model includes M feature layers and a task layer (M≥1). Each feature layer of each task model includes a parameter matrix that is different for the different models. An encoder-decoder model is trained. The encoder-decoder model includes sequentially: an input layer, an encoder, M hidden layers, a decoder, and an output layer. The encoder is a neural network that maps and compresses the parameter matrices in the input layer into the M hidden layers, which generates a basic state model. The decoder is a neural network that receives the basic state model as input and generates the output layer to be identical to the input layer.

    INTENTION IDENTIFICATION IN DIALOGUE SYSTEM
    5.
    发明申请
    INTENTION IDENTIFICATION IN DIALOGUE SYSTEM 有权

    公开(公告)号：US20230127907A1

    公开(公告)日：2023-04-27

    申请号：US17451836

    申请日：2021-10-22

    申请人： INTERNATIONAL BUSINESS MACHINES CORPORATION

    发明人： Zhong Fang Yuan ,  Tong Liu ,  Li Juan Gao ,  Yi Chen Zhong ,  Hai Bo Zou

    IPC分类号： G06F40/35 ,  G06F40/295 ,  G06N5/00

    摘要： Embodiments of the present disclosure relate to question answering. A computer-implemented method includes determining a plurality of intention candidates of a user from the user's question; determining a set of entities and attributes associated with the set of entities from the plurality of intention candidates; constructing a decision tree from the set of entities and the attributes associated with the set of entities, wherein each node of the decision tree is associated with a respective one of the attributes and represents a respective subset of the plurality of intention candidates, and wherein the respective subset of the plurality of intention candidates are split based on the entities associated with the respective one of the attributes; and generating a question corresponding to a node of the decision tree to determine the user's intention.

    RAPID LANGUAGE DETECTION FOR CHARACTERS IN IMAGES OF DOCUMENTS
    6.
    发明申请
    RAPID LANGUAGE DETECTION FOR CHARACTERS IN IMAGES OF DOCUMENTS 有权

    公开(公告)号：US20230073932A1

    公开(公告)日：2023-03-09

    申请号：US17468474

    申请日：2021-09-07

    申请人： International Business Machines Corporation

    发明人： Zhong Fang Yuan ,  Tong Liu ,  Li Juan Gao ,  Xiang Yu Yang ,  Qiang He ,  Yu Pan

    IPC分类号： G06F40/279 ,  G06K9/00 ,  G06N3/08 ,  G06N20/10

    摘要： A computer-implemented method, according to one embodiment, includes: receiving an image having characters that correspond to a language, and using a text recognition algorithm to determine a first language believed to correspond to the characters. A first confidence level associated with the first language is also computed, and a determination is made as to whether the first confidence level associated with the first language is outside a predetermined range. In response to determining that the first confidence level associated with the first language is not outside the predetermined range, the first language is output as the given language. The text recognition algorithm is trained using a simple shallow neural network and a generated mixed language corpus. The generated mixed language corpus is formed by: randomly sampling libraries having vocabulary and/or characters therein, and combining the randomly sampled vocabulary and/or characters to form the generated mixed language corpus.

    DATA-DRIVEN NEURAL NETWORK MODEL COMPRESSION
    9.
    发明申请
    DATA-DRIVEN NEURAL NETWORK MODEL COMPRESSION 有权

    公开(公告)号：US20220180180A1

    公开(公告)日：2022-06-09

    申请号：US17115857

    申请日：2020-12-09

    申请人： International Business Machines Corporation

    发明人： Tong Liu ,  Zhong Fang Yuan ,  Kun Yan Yin ,  He Li ,  Li Juan Gao

    IPC分类号： G06N3/08 ,  G06N3/04

    摘要： A data-driven model compression technique is introduced that only targets to provide same accuracy as the original (not compressed) model in certain areas by reducing compression parameters. A compression engine relies on backpropagation to determine an extent of parameter value changes and designate certain parameters as key parameters. The model matrix is reshaped according to importance of each neuron. Only randomly generated parameter values of the reshaped parameter matrix are fine tuned to create a reliable compressed neural network model.