Patents by Inventor Xiangxiang CHU

Xiangxiang CHU has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Patent number: 11663468
    Abstract: A method for training a neural network, includes: training a super network to obtain a network parameter of the super network, wherein each network layer of the super network includes multiple candidate network sub-structures in parallel; for each network layer of the super network, selecting, from the multiple candidate network sub-structures, a candidate network sub-structure to be a target network sub-structure; constructing a sub-network based on target network sub-structures each selected in a respective network layer of the super network; and training the sub-network, by taking the network parameter inherited from the super network as an initial parameter of the sub-network, to obtain a network parameter of the sub-network.
    Type: Grant
    Filed: January 16, 2020
    Date of Patent: May 30, 2023
    Assignee: Beijing Xiaomi Intelligent Technology Co., Ltd.
    Inventors: Xiangxiang Chu, Ruijun Xu, Bo Zhang, Jixiang Li, Qingyuan Li, Bin Wang
  • Patent number: 11580408
    Abstract: A search method for a neural network model structure, includes: generating an initial generation population of network model structure based on multi-objective optimization hyper parameters, as a current generation population of network model structure; performing selection and crossover on the current generation population of network model structure; generating a part of network model structure based on reinforcement learning mutation, and generating a remaining part of network model structure based on random mutation on the selected and crossed network model structure; generating a new population of network model structure based on the part of network model structure generated by reinforcement learning mutation and the remaining part of network model structure generated by random mutation; and searching a next generation population of network model structure based on the current generation population of network model structure and the new population of network model structure.
    Type: Grant
    Filed: March 26, 2020
    Date of Patent: February 14, 2023
    Assignee: Beijing Xiaomi Intelligent Technology Co., Ltd.
    Inventors: Xiangxiang Chu, Ruijun Xu, Bo Zhang, Jixiang Li, Qingyuan Li
  • Patent number: 11443189
    Abstract: A hypernetwork training method includes: acquiring a multipath neural subnetwork based on a preconstructed initial hypernetwork; training the multipath neural subnetwork to update a weight parameter of each substructure in the multipath neural subnetwork; synchronizing the weight parameter of each substructure in the multipath neural subnetwork to the preconstructed initial hypernetwork; and determining whether the preconstructed initial hypernetwork converges, and if it is determined that the preconstructed initial hypernetwork does not converge, re-executing the acquiring, the training, the synchronizing, and the determining, to obtain a target hypernetwork.
    Type: Grant
    Filed: March 24, 2020
    Date of Patent: September 13, 2022
    Assignee: Beijing Xiaomi Intelligent Technology Co., Ltd.
    Inventors: Xiangxiang Chu, Bo Zhang, Ruijun Xu, Bin Wang
  • Publication number: 20210390449
    Abstract: Provided are a method and device for data processing and a storage medium. The method includes: inputting labeled training data in a training set into a preset model to be trained and updating the model parameters of the preset model; inputting labeled verification data in a verification set into the preset model after the model parameters are updated to obtain a first prediction label; obtaining a verification loss value based on a difference between the first prediction label and a marked label of the labeled verification data; determining an auxiliary loss value based on current structural parameters of the preset model; determining whether to stop training the preset model based on the verification loss value and the auxiliary loss value; and classifying data to be classified based on a target network model constructed by a network structure included in the trained preset model to obtain a classification result.
    Type: Application
    Filed: January 12, 2021
    Publication date: December 16, 2021
    Inventors: Xiangxiang Chu, Bo Zhang, Tianbao Zhou, Bin Wang
  • Patent number: 11189014
    Abstract: A method for processing an image includes: an image to be processed with a first resolution is acquired; and the image to be processed is processed by a target neural network model to obtain a target image, the target image being a denoised image with a second resolution, the second resolution being higher than the first resolution, and the target neural network model including a first preset number of convolutional layers and a second preset number of sub-pixel up-sampling portions.
    Type: Grant
    Filed: November 18, 2019
    Date of Patent: November 30, 2021
    Assignee: BEIJING XIAOMI MOBILE SOFTWARE CO., LTD.
    Inventors: Hailong Ma, Xiangxiang Chu, Qike Zhao
  • Publication number: 20210334661
    Abstract: The present disclosure relates to an image processing method and apparatus based on a super network, and a computer storage medium. The method can include that a pretrained backbone network is merged with a rear end of a target detection network to obtain a merged super network, the merged super network is trained, Neural Architecture Search (NAS) is performed based on the trained super network to obtain a target detection neural architecture, and an image to be processed is processed by using the target detection neural architecture to obtain an image processing result.
    Type: Application
    Filed: September 22, 2020
    Publication date: October 28, 2021
    Applicant: Beijing Xiaomi Pinecone Electronics Co., Ltd.
    Inventors: Xiangxiang CHU, Ruijun XU, Bo ZHANG, Bin WANG
  • Patent number: 11151692
    Abstract: A method for processing an image includes: an image to be processed with a first resolution is acquired; and the image to be processed is processed by a target neural network model to obtain a target image, the target image being a denoised image with a second resolution, the second resolution being higher than the first resolution, and the target neural network model including a first preset number of convolutional layers and a second preset number of sub-pixel up-sampling portions.
    Type: Grant
    Filed: November 18, 2019
    Date of Patent: October 19, 2021
    Assignee: BEIJING XIAOMI MOBILE SOFTWARE CO., LTD.
    Inventors: Hailong Ma, Xiangxiang Chu, Qike Zhao
  • Publication number: 20210142166
    Abstract: A hypernetwork training method includes: acquiring a multipath neural subnetwork based on a preconstructed initial hypernetwork; training the multipath neural subnetwork to update a weight parameter of each substructure in the multipath neural subnetwork; synchronizing the weight parameter of each substructure in the multipath neural subnetwork to the preconstructed initial hypernetwork; and determining whether the preconstructed initial hypernetwork converges, and if it is determined that the preconstructed initial hypernetwork does not converge, re-executing the acquiring, the training, the synchronizing, and the determining, to obtain a target hypernetwork.
    Type: Application
    Filed: March 24, 2020
    Publication date: May 13, 2021
    Inventors: Xiangxiang Chu, Bo Zhang, Ruijun Xu, Bin Wang
  • Publication number: 20210133563
    Abstract: A method for training a neural network, includes: training a super network to obtain a network parameter of the super network, wherein each network layer of the super network includes multiple candidate network sub-structures in parallel; for each network layer of the super network, selecting, from the multiple candidate network sub-structures, a candidate network sub-structure to be a target network sub-structure; constructing a sub-network based on target network sub-structures each selected in a respective network layer of the super network; and training the sub-network, by taking the network parameter inherited from the super network as an initial parameter of the sub-network, to obtain a network parameter of the sub-network.
    Type: Application
    Filed: January 16, 2020
    Publication date: May 6, 2021
    Inventors: Xiangxiang Chu, Ruijun Xu, Bo Zhang, Jixiang Li, Qingyuan Li, Bin Wang
  • Publication number: 20210110276
    Abstract: A search method for a neural network model structure, includes: generating an initial generation population of network model structure based on multi-objective optimization hyper parameters, as a current generation population of network model structure; performing selection and crossover on the current generation population of network model structure; generating a part of network model structure based on reinforcement learning mutation, and generating a remaining part of network model structure based on random mutation on the selected and crossed network model structure; generating a new population of network model structure based on the part of network model structure generated by reinforcement learning mutation and the remaining part of network model structure generated by random mutation; and searching a next generation population of network model structure based on the current generation population of network model structure and the new population of network model structure.
    Type: Application
    Filed: March 26, 2020
    Publication date: April 15, 2021
    Inventors: Xiangxiang CHU, Ruijun XU, Bo ZHANG, Jixiang LI, Qingyuan LI
  • Publication number: 20210065004
    Abstract: A method for subnetwork sampling is applicable to a hypernetwork topoloyly. The hypernetwork topology includes n layers, each layer includes at least two substructures, and each substructure includes hatch normalization (BN) modules in one-to-one correspondence with the substructures of a closest upper layer, n>0 and n being a positive integer. The method includes: a substructure A(N) of an N-th layer is selected, 1>N?n; a selected substructure A(N-1) of an (N?1)-th layer is determined; a BN module C(B) in one-to-to correspondence with A(N-1) is determined from the substructure A(N); and the substructure A(N) is added into a subnetwork through the BN module C(B).
    Type: Application
    Filed: November 20, 2019
    Publication date: March 4, 2021
    Inventors: Xiangxiang CHU, Ruijun XU, Bo ZHANG, Jixiang LI, Qingyuan LI, Bin WANG
  • Publication number: 20210056421
    Abstract: A supernet construction method includes: setting a linear connection unit in at least one layer of a supernet, wherein an input end of the linear connection unit is connected to an upper layer of a home layer of the linear connection unit, and an output end is connected to a lower layer of the home layer of the linear connection unit; an output and an input of the linear connection unit form a linear relationship, where the linear relationship includes a linear relationship other than that the output is equal to the input.
    Type: Application
    Filed: November 28, 2019
    Publication date: February 25, 2021
    Applicant: Beijing Xiaomi Intelligent Technology Co., Ltd.
    Inventors: Xiangxiang CHU, Ruijun XU, Bo ZHANG, Jixiang LI, Qingyuan LI, Bin WANG
  • Publication number: 20210027426
    Abstract: A method for processing an image includes: an image to be processed with a first resolution is acquired; and the image to be processed is processed by a target neural network model to obtain a target image, the target image being a denoised image with a second resolution, the second resolution being higher than the first resolution, and the target neural network model including a first preset number of convolutional layers and a second preset number of sub-pixel up-sampling portions.
    Type: Application
    Filed: November 18, 2019
    Publication date: January 28, 2021
    Applicant: BEIJING XIAOMI MOBILE SOFTWARE CO., LTD.
    Inventors: Hailong MA, Xiangxiang CHU, Qike ZHAO
  • Publication number: 20200387795
    Abstract: A super network training method includes: performing sub-network sampling on a super network for multiple rounds to obtain a plurality of sub-networks, wherein for any layer of the super network, different sub-structures are selected when sampling different sub-networks, and training the plurality of sub-networks obtained by sampling and updating the super network.
    Type: Application
    Filed: November 25, 2019
    Publication date: December 10, 2020
    Inventors: Xiangxiang CHU, Ruijun XU, Bo ZHANG, Jixiang LI, Qingyuan LI