PerMo (PersonaMotion)
收藏资源简介:
PerMo是一个由多位演员捕捉的独特个性化动作的新大规模数据集,由伯明翰大学等研究机构创建。该数据集包含多种类型的动作,旨在支持动作个性化任务。通过特定的捕捉技术和数据处理方法,PerMo能够捕捉到每位演员独特的个性化特征。数据集的应用领域主要是虚拟现实和元宇宙中的动作生成,旨在解决如何通过文本描述生成与个人特质相匹配的独特动作的问题。
PerMo is a novel large-scale dataset that captures unique personalized motions from multiple actors, developed by research institutions including the University of Birmingham. This dataset includes a diverse range of motion types, aiming to support motion personalization tasks. By adopting specialized motion capture technologies and data processing methods, PerMo is able to capture the distinct personalized traits of each actor. The primary application areas of this dataset are motion generation in virtual reality (VR) and the metaverse, where it addresses the challenge of generating unique motions that align with an individual's inherent traits based on textual descriptions.




