公开(公告)号：US20160248527A1

公开(公告)日：2016-08-25

申请号：US14707426

申请日：2015-05-08

申请人： INSTITUTE FOR INFORMATION INDUSTRY

发明人： Yao-Min Huang ,  Wen-Shan Liou

IPC分类号： H04H60/66 ,  G10L25/72 ,  H04H60/59 ,  H04H60/58 ,  H04N21/442 ,  G10L15/30 ,  H04N21/24

CPC分类号： H04H60/66 ,  H04H60/372 ,  H04H60/40 ,  H04H60/42 ,  H04H60/44 ,  H04H60/48 ,  H04H60/65 ,  H04N21/4394 ,  H04N21/44008

摘要： A distributed multimedia recognition system comprises a user equipment and a server platform, wherein the user equipment is connected to the server platform. The user equipment transmits a recognition request, and performs characteristic extraction on a multimedia, so as to obtain characteristic data. The server platform receives the recognition request, selects one or multiple batches of corresponding reference characteristic data among multiple batches of reference characteristic data recorded in a characteristic database thereof according to a condition of the recognition request, and sends the one or multiple batches of the corresponding reference characteristic data to the user equipment. The user equipment compares the characteristic data to the one or multiple batches of the corresponding reference characteristic data to obtain a recognition result.

摘要翻译： 分布式多媒体识别系统包括用户设备和服务器平台，其中用户设备连接到服务器平台。 用户设备发送识别请求，并对多媒体进行特征提取，以获得特征数据。 服务器平台接收识别请求，根据识别请求的条件，在记录在其特征数据库中的多批参考特征数据之间选择一批或多批对应的参考特征数据，并发送对应的一批或多批 参考特征数据给用户设备。 用户设备将特征数据与相应参考特征数据的一批或多批进行比较，以获得识别结果。

公开(公告)号：US10212291B2

公开(公告)日：2019-02-19

申请号：US15673407

申请日：2017-08-09

申请人： INSTITUTE FOR INFORMATION INDUSTRY

发明人： Yao-Min Huang ,  Wen-Shan Liou ,  Hsin-I Lai

IPC分类号： H04N5/228 ,  H04N1/00 ,  G06K9/46

摘要： A recognition system includes an image capturing device and a server. The image capturing device generates a Mth-layer calculation result based on image data and a convolutional neural network (CNN), and transmits feature information associated with the Mth-layer calculation result. M is a positive integer, M is equal to or greater than 1, M is less than or equal to N, and N is a predetermined positive integer. The server receives the feature information. The server generates a Kth-layer calculation result based on the feature information and the CNN by an iterative method when M is less than N. K is a positive integer which is greater than M and less than or equal to N. The server generates a first recognition result associated with the image data based on the Kth-layer calculation result and a first recognition model when K is equal to N.

公开(公告)号：US12008688B2

公开(公告)日：2024-06-11

申请号：US17446457

申请日：2021-08-30

申请人： INSTITUTE FOR INFORMATION INDUSTRY

发明人： Kai-Hsuan Chan ,  Yao-Min Huang

IPC分类号： G06T11/00 ,  A61B6/02 ,  A61B6/50 ,  G06F18/213 ,  G06N3/04 ,  G06T7/00 ,  G16H30/20 ,  G16H30/40 ,  G16H50/20

CPC分类号： G06T11/003 ,  A61B6/025 ,  A61B6/502 ,  G06F18/213 ,  G06N3/04 ,  G06T7/0012 ,  G16H30/20 ,  G16H30/40 ,  G16H50/20 ,  G06T2207/10112 ,  G06T2207/20081 ,  G06T2207/20084 ,  G06T2207/30068

摘要： A method for processing breast tomosynthesis images includes: obtaining multiple breast images, performing binarization operation and pooling operation to at least one marked breast image to obtain a lesion mask; training a convolutional neural network (CNN) according to the breast images and the lesion mask; obtaining multiple cropped images from the breast images of different layers according to a particular two-dimensional position, and inputting each cropped image into the CNN to obtain cropped heat maps; inputting the cropped heat maps into a recurrent neural network (RNN) in an order of the levels to obtain output heat maps, and combining the output heat maps to obtain a integrated heat map, and training the RNN according to the integrated heat map and the lesion mask.

公开(公告)号：US09165067B2

公开(公告)日：2015-10-20

申请号：US14174857

申请日：2014-02-07

申请人： INSTITUTE FOR INFORMATION INDUSTRY

发明人： Chung-Che Wang ,  Yao-Min Huang ,  Pei-Yu Liao

IPC分类号： G06F17/30

CPC分类号： G06F17/3074 ,  G06F17/30743

摘要： A computer system is configured for performing audio matching. The computer system includes a client terminal and a server. The client terminal includes a classifier, a first landmark extraction module, a first encoder module and an integration module. The classifier is configured for estimating an identification difficulty. The first landmark extraction module is configured for determining the number of times for landmark extraction, and for extracting first landmarks. The first encoder module is configured for encoding the first landmarks. The server includes a second landmark extraction module, a second encoder module, a hash table, a table-lookup module, an analysis module and a decision module. The second landmark extraction module is configured for extracting second landmarks. The second encoder module is configured for encoding the second landmarks. The table-lookup module is configured for performing a table lookup. The decision module is configured for determining an audio matching result.

摘要翻译： 计算机系统被配置为执行音频匹配。 计算机系统包括客户终端和服务器。 客户终端包括分类器，第一地标提取模块，第一编码器模块和集成模块。 分类器被配置为估计识别难度。 第一地标提取模块被配置用于确定地标提取的次数，以及用于提取第一地标。 第一编码器模块被配置用于对第一地标进行编码。 服务器包括第二地标提取模块，第二编码器模块，散列表，表查找模块，分析模块和决策模块。 第二地标提取模块被配置用于提取第二地标。 第二编码器模块被配置为对第二地标进行编码。 表查找模块被配置为执行表查找。 决定模块被配置用于确定音频匹配结果。