• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>ACM Transactions on Graphics >Generating and Ranking Diverse Multi-Character Interactions
【24h】

Generating and Ranking Diverse Multi-Character Interactions

机译：生成和排序不同的多字符交互

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

In many application areas, such as animation for pre-visualizingrnmovie sequences and choreography for dance or other types of performance,rnonly a high-level description of the desired scene is providedrnas input, either written or verbal. Such sparsity, however,rnlends itself well to the creative process, as the choreographer, animatorrnor director can be given more choice and control of the finalrnscene. Animating scenes with multi-character interactions can berna particularly complex process, as there are many different constraintsrnto enforce and actions to synchronize. Our novel ‘generateand-rnrank’ approach rapidly and semi-automatically generates datadrivenrnmulti-character interaction scenes from high-level graphicalrndescriptions composed of simple clauses and phrases. From arndatabase of captured motions, we generate a multitude of plausiblerncandidate scenes. We then efficiently and intelligently rank thesernscenes in order to recommend a small but high-quality and diversernselection to the user. This set can then be refined by re-ranking or byrngenerating alternatives to specific interactions. While our approachrnis applicable to any scenes that depict multi-character interactions,rnwe demonstrate its efficacy for choreographing fighting scenes andrnevaluate it in terms of performance and the diversity and coveragernof the results.
机译：在许多应用领域中，例如用于预先可视化电影序列的动画和用于舞蹈或其他类型表演的编排，仅提供对期望场景的高级描述，无论是书面还是口头输入。但是，这种稀疏性很适合于创作过程，因为可以为编舞者，动画导演提供更多选择和控制最终场景。具有多角色交互的动画场景可能会造成特别复杂的过程，因为要执行的约束和要同步的动作有很多不同。我们新颖的“生成和排列”方法可以根据由简单的从句和短语组成的高级图形描述快速，半自动地生成数据驱动的多字符交互场景。从捕获的运动的arndatabase中，我们生成了许多合理的候选场景。然后，我们可以高效，智能地对场景进行排名，以向用户推荐一个小而高质量的多样化选择。然后可以通过重新排序或通过生成特定交互的替代项来完善此集合。尽管我们的方法适用于描述多角色交互的任何场景，但我们展示了其对战斗场景进行编排的功效，并根据性能以及结果的多样性和覆盖范围对其进行了重新评估。

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