公开(公告)号：US10474925B2

公开(公告)日：2019-11-12

申请号：US15793086

申请日：2017-10-25

Applicant: INDUSTRIAL TECHNOLOGY RESEARCH INSTITUTE

Inventor： Mao-Yu Huang ,  Ching-Hao Lai

IPC: G06K9/62 ,  G06N3/02

Abstract: A deep neural network and a method for recognizing and classifying a multimedia data as one of a plurality of pre-determined data classes with enhanced recognition and classification accuracy and efficiency are provided. The use of the side branch(es) (or sub-side branch(es), sub-sub-side branch(es), and so on) extending from the main branch (or side branch(es), sub-side branch(es), and so on), the sequential decision making mechanism, and the collaborating (fusing) decision making mechanism in a deep neural network would equip a deep neural network with the capability for fast forward inference so as to enhance recognition and classification accuracy and efficiency of the deep neural network.

公开(公告)号：US11704534B2

公开(公告)日：2023-07-18

申请号：US16221612

申请日：2018-12-17

Applicant: Industrial Technology Research Institute

Inventor： Ching-Hao Lai ,  Mao-Yu Huang

IPC: G06F18/245 ,  G06F18/2113 ,  G06N3/02 ,  G06N3/04

CPC classification number: G06N3/02 ,  G06F18/2113 ,  G06F18/245 ,  G06N3/04

Abstract: Provided is a neural-network-based classification method, including: generating, by a neural network, one or more score vectors corresponding to one or more samples respectively; determining a first subset of the one or more samples according to the one or more score vectors and a first decision threshold, wherein the first subset is associated with a first class; and selecting samples to be re-examined from the one or more samples according to the first subset.

公开(公告)号：US11547350B2

公开(公告)日：2023-01-10

申请号：US16232400

申请日：2018-12-26

Applicant: INDUSTRIAL TECHNOLOGY RESEARCH INSTITUTE

Inventor： Chih-Tsung Shen ,  Hsin-Jung Cheng ,  Ching-Hao Lai ,  Szu-Han Tzao

IPC: A61B5/00 ,  A61M21/02 ,  A61M21/00

Abstract: A personalized parameter learning method, a sleep-aid device and a non-transitory computer readable medium are provided. The personalized parameter learning method for a sleep-aid device is provided. The personalized parameter learning method includes the following steps. A process device computes a measured sleep quality of a user after operating a sleep-aid device with an inputted parameter setting at least according to a subjective feedback from the user. The processing device generates a plurality of candidate parameter settings according to the measured sleep quality. The processing device generates a plurality of predicting sleep qualities corresponding the candidate parameter settings. The processing device obtains a recommending parameter setting by selecting one of the candidate parameter settings according to the predicting sleep qualities.