首页> 外国专利> DANCE TRAINING APPARATUS AND METHOD USING AUTOMATIC GENERATION OF DANCE KEY MOTION

DANCE TRAINING APPARATUS AND METHOD USING AUTOMATIC GENERATION OF DANCE KEY MOTION

机译:舞动动作自动生成的舞动训练装置及方法

摘要

Disclosed are dance training apparatus and method using automatic generation of dance key motions. The dance training apparatus includes: a generation unit (10) for generating standard movement information according to a time change by recognizing movement from a dance video; a filming unit (20) for recognizing movement from a video obtained after filming a user and generating comparative movement information; and an output unit (30) for calculating a difference between the standard movement information and the comparative movement information and outputting a congruence level. Therefore, the present invention can output a congruence level after comparing movement of a dance video with the user′s movement and play a next dance video corresponding to an excess section so that the user can undergo intensive training with respect to the section exceeding a difference between the movement of the dance video with the user′s movement, can generate a dance video using a plurality of characters included in the dance video and reflect voices, facial expressions and gestures according to features of the characters, can read the dance video from digital media to analyze images and infer standard movements according to a time change, can extract the user′s movement video from the video filming the user and add it to the dance video, can recognize movement by applying joint links to the user′s body area and recognize movement by applying a neural network, and can calculate dance difficulty level by collecting other users′ congruence levels, and make and output scores on the congruence level by reflecting the dance difficulty level.
机译:公开了一种使用自动产生舞蹈关键动作的舞蹈训练设备和方法。该舞蹈训练设备包括:生成单元(10),用于通过根据舞蹈视频识别运动来根据时间变化来生成标准运动信息;以及拍摄单元(20),用于从拍摄用户后获得的视频中识别运动并生成比较运动信息;输出单元(30),其用于计算标准运动信息和比较运动信息之间的差并输出同等水平。因此,本发明可以在将舞蹈视频的运动与用户的运动进行比较之后输出一致度,并播放与多余部分相对应的下一舞蹈视频,从而使得用户可以对超过差异的部分进行强化训练。在舞蹈视频的运动与用户的运动之间,可以使用舞蹈视频中包括的多个字符生成舞蹈视频,并根据角色的特征反映语音,面部表情和手势,可以从中读取舞蹈视频。数字媒体可以分析图像并根据时间变化推断标准动作,可以从拍摄用户的视频中提取用户的动作视频并将其添加到舞蹈视频中,可以通过将链接应用于用户的身体来识别动作通过应用神经网络进行区域划分和识别动作,并可以通过收集其他用户的同等程度来计算舞蹈难度等级,并在他通过反映舞蹈难度水平来达到同等水平。

著录项

  • 公开/公告号KR101936692B1

    专利类型

  • 公开/公告日2019-01-10

    原文格式PDF

  • 申请/专利权人 SOGWARE;

    申请/专利号KR20170173931

  • 发明设计人 SOHN KANG MIN;

    申请日2017-12-18

  • 分类号A63F13/213;A63B69;G06K9;G06T7/20;

  • 国家 KR

  • 入库时间 2022-08-21 11:49:22

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

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