• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Seminar on Application for Technology of Information and Communication >Classification of Music Moods Based on CNN
【24h】

Classification of Music Moods Based on CNN

机译：基于CNN的音乐情绪分类

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

In recent years, music has become part of everyone life. Almost everyone with different backgrounds listening to music, as a form of entertainment or just accompany in doing the activity. Music can be enjoyed in various ways based on the level of society. Usually music is categorized by genre, which shows the type of music or the form of music. Like pop, rock, metal, ballad, dangdut, jazz, keroncong, and etc. But the division by genre is common. So, in this study we try to classify a music into some moods like happy, peaceful, angry, and sad. Because such a classification will be easily accepted by listeners, especially people who are not very familiar with the genre. This classification is performed with 200 track databases for the training data and 50 track database for testing. We use a music features, such as pitch, pulse clarity, tempo, key, and scale. For the classification method in this paper we used Convolutional Neural Networks (CNN), the accuracy that we have are 82%.
机译：近年来，音乐已成为每个人生的一部分。几乎每个人都有不同的背景，听音乐，作为一种娱乐形式，或者只是陪伴活动。基于社会水平，可以以各种方式享受音乐。通常音乐由类型分类，显示音乐类型或音乐的形式。像流行音乐，岩石，金属，民谣，唐丁，爵士，克罗尼克等。但是划分是常见的。因此，在这项研究中，我们试图将音乐分类为一些情绪，如快乐，和平，生气和悲伤。因为这种分类将很容易被倾听者接受，特别是那些对流派不太熟悉的人。使用200个轨道数据库执行此分类，用于训练数据和50个跟踪数据库进行测试。我们使用音乐功能，例如音高，脉冲清晰度，节奏，键和缩放。对于本文的分类方法，我们使用了卷积神经网络（CNN），我们拥有的准确性为82％。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号