EmoAva
收藏资源简介:
EmoAva是由哈尔滨工业大学(深圳)等机构创建的一个大规模、高质量的文本到3D表情映射数据集,包含15,000个实例。该数据集旨在解决从文本生成情感丰富的3D面部表情的问题,通过提供多样化的表情、情感内容一致性和表情流畅性的映射,推动情感计算和3D虚拟形象生成领域的发展。数据集的创建过程结合了先进的3D面部模型FLAME,并通过一系列评估指标确保模型的有效性。EmoAva的应用领域广泛,包括虚拟客服、在线治疗等需要情感表达的场景。
EmoAva is a large-scale and high-quality text-to-3D facial expression mapping dataset created by institutions such as Harbin Institute of Technology (Shenzhen). Comprising 15,000 instances, this dataset is designed to address the challenge of generating emotionally rich 3D facial expressions from text. By providing diverse expressions, consistent emotional content, and fluid expression mapping, it promotes the development of affective computing and 3D virtual avatar generation. The dataset's creation process integrates the advanced 3D facial model FLAME and ensures the effectiveness of the model through a series of evaluation metrics. EmoAva has a wide range of applications, including scenarios requiring emotional expression such as virtual customer service and online therapy.

- 1Towards Rich Emotions in 3D Avatars: A Text-to-3D Avatar Generation Benchmark哈尔滨工业大学(深圳) · 2024年



