The Best Practice of University and Community Cooperation in Open Source Software Project - TV Station Media Images Query System for Example

被引：3

作者：

Lin, Feng-Cheng ^{[1
]}

Chen, Hsi-Min ^{[1
]}

Lin, Shih-Feng ^{[1
]}

Chu, Hsing-Yi ^{[2
]}

机构：

[1] Feng Chia Univ, Dept Informat Engn & Comp Sci, Taichung, Taiwan

[2] New Media Dept, Publ Televis Serv, Taipei, Taiwan

来源：

2018 15TH INTERNATIONAL SYMPOSIUM ON PERVASIVE SYSTEMS, ALGORITHMS AND NETWORKS (I-SPAN 2018) | 2018年

关键词：

Images Query System; Open Source Project; Deep Learning; Convolutional Neural Networks;

D O I：

10.1109/I-SPAN.2018.00065

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

One open source software project, Open Source Technology Development and Cultivation of Talent, of Institute for Information Industry (III), Taiwan released an opportunity and gave us a valuable practice in helping some engineers, students, and teachers to work together. In this project, we try to develop the TV station media images query system prototype. Face recognition based on videos or image sets has been involved in the project. We try to apply Convolutional Neural Networks (CNN) and perform feature extraction on a target face in multiple image frames of given videos and generate multiple face feature vectors respectively. This proposed prototype and flowchart try to convert the plurality of face feature vectors into a feature vector of a predetermined dimension and judge the feature vector of the predetermined dimension by using a classifier to recognize the target face. The user can input Chinese name of actress and pressure search button, then, if information retrieved from the database associated with a selected person, it can show all similar images and time associated with a given face name.

引用

页码：346 / 349

页数：4