Patents by Inventor Shih-Jen Chu

Shih-Jen 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).

  • Publication number: 20240013523
    Abstract: A model training method and a model training system are disclosed. The method includes the following. In a first iteration training, first training data corresponding to a first sub-task and second training data corresponding to a second sub-task are input to a target model. A first and a second image identification rate of the target model with respect to the first and the second sub-task are evaluated respectively according to a first output corresponding to the first sub-task and a second output corresponding to the second sub-task from the target model. A first and a second sampling rate respectively corresponding to the first and the second training data in a second iteration training are adjusted according to the first and the second image identification rate. The first and the second sampling rate are respectively negatively correlated to the first and the second image identification rate.
    Type: Application
    Filed: May 5, 2023
    Publication date: January 11, 2024
    Applicant: PEGATRON CORPORATION
    Inventor: Shih-Jen Chu
  • Patent number: 11756179
    Abstract: A training method for an anomaly detection model and an electronic device using the same are provided. The anomaly detection model includes a generative model and a discriminative model. The training method for the anomaly detection model includes the following steps. One of a plurality of original images and one of a plurality of task information are used as a training sample. The training sample is input to the generative model and the discriminative model to calculate a plurality of network loss results corresponding to the training sample. If the original image of the training sample does not match the task information of the training sample, a first loss function is obtained based on a weighted sum of reciprocals of the network loss results, and the generative model is trained according to the first loss function.
    Type: Grant
    Filed: October 27, 2020
    Date of Patent: September 12, 2023
    Assignee: PEGATRON CORPORATION
    Inventor: Shih-Jen Chu
  • Publication number: 20220327381
    Abstract: A model training apparatus, a model training method, and a computer-readable medium are provided. In the method, a labeled abnormal sample is inputted into an abnormal detecting model. The abnormal detecting model is based on an autoencoder structure. A reconstructed error between the abnormal sample and an output of the abnormal detecting model is maximized to optimize the abnormal detecting model.
    Type: Application
    Filed: February 10, 2022
    Publication date: October 13, 2022
    Applicant: PEGATRON CORPORATION
    Inventor: Shih-Jen Chu
  • Publication number: 20220067583
    Abstract: A method and an electronic device for evaluating a performance of an identification model are provided. The method includes: obtaining a source data sample, a plurality of test samples, and a target data sample; inputting the plurality of test samples into a pre-trained model trained based on the source data sample to obtain a normal sample and an abnormal sample; converting the source data sample to generate a converted source data sample, converting the normal sample to generate a converted normal sample, and converting the abnormal sample to generate a converted abnormal sample; adjusting the pre-trained model to obtain the identification model according to the converted source data sample and the target data sample; and inputting the converted normal sample and the converted abnormal sample into the identification model to evaluate the performance of the identification model.
    Type: Application
    Filed: July 6, 2021
    Publication date: March 3, 2022
    Applicant: PEGATRON CORPORATION
    Inventor: Shih-Jen Chu
  • Publication number: 20210150698
    Abstract: A training method for an anomaly detection model and an electronic device using the same are provided. The anomaly detection model includes a generative model and a discriminative model. The training method for the anomaly detection model includes the following steps. One of a plurality of original images and one of a plurality of task information are used as a training sample. The training sample is input to the generative model and the discriminative model to calculate a plurality of network loss results corresponding to the training sample. If the original image of the training sample does not match the task information of the training sample, a first loss function is obtained based on a weighted sum of reciprocals of the network loss results, and the generative model is trained according to the first loss function.
    Type: Application
    Filed: October 27, 2020
    Publication date: May 20, 2021
    Applicant: PEGATRON CORPORATION
    Inventor: Shih-Jen Chu