The Visualization System of Image Search Base on Convolutional Spherical Self Organizing Map Implemented Using WebGL
The Visualization System of Image Search Base on Convolutional Spherical Self Organizing Map Implemented Using WebGL
复制标题
利用WebGL实现的基于卷积球自组织图的图像搜索可视化系统
DOI：
10.1007/978-981-16-2102-4_46
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
and Gen Niina
中科院分区：
文献类型：
--
作者：
Hiroshi Dozono;Kenta Toyozumi;Koki Yoshioka;and Gen Niina
最近，从小型物联网设备到大型计算机系统的许多数字设备创建了大量数据。为了帮助人类理解，重要的是可视化数据之间的关系。作为方法之一，卷积球面自组织映射（卷积球面SOM）在这项研究中提出。此外，使用JavaScript和WebGL来实现球形SOM的3D绘制和操作，因此可以使用Web浏览器在许多设备上执行。该方法的有效性得到了证实，通过实现一个系统，可以可视化从互联网上搜索的球面SOM的图像之间的关系。
Recently, many digital devices from small IoT devices to large systems of computers create a huge number of data. To help the understanding by human, it is important to visualize the relations among the data. As one of a method, Convolutional Spherical Self Organizing Map (Convolutional Spherical SOM) is proposed in this research. Additionally, Javascript and WebGL are used to implement the 3D drawing and manipulation of Spherical SOM, thus it becomes possible to execute on many devices using Web browsers. The effectiveness of this method is confirmed by implementing a system that can visualize the relations among the images searched from the Internet on Spherical SOM.