论文标题
如果可以的话，请描述我！特征是实例级人类解析
Describe me if you can! Characterized Instance-level Human Parsing
论文作者
论文摘要
人类搜索或在线时尚等几种计算机视觉应用都取决于人类描述。因此，实例级人类解析（HP）的使用非常重要，因为它将语义属性和身体位置在一个人内部。但是如何表征这些属性？据我们所知，只有一些单HP数据集描述具有某些颜色，大小和/或模式特征的属性。在野外缺乏具有此类特征的多HP数据集。在本文中，我们根据多HP数据集CIHP提出数据集CCIHP，其中20个新标签涵盖了这三种特征。此外，我们提出了HPTR，这是一种基于变压器的新自下而上的多任务方法，作为快速可扩展的基线。这是多HP最快的最快方法，同时具有与最精确的自下而上方法相当的精确度。我们希望这将鼓励研究快速准确的精确人类描述方法。
Several computer vision applications such as person search or online fashion rely on human description. The use of instance-level human parsing (HP) is therefore relevant since it localizes semantic attributes and body parts within a person. But how to characterize these attributes? To our knowledge, only some single-HP datasets describe attributes with some color, size and/or pattern characteristics. There is a lack of dataset for multi-HP in the wild with such characteristics. In this article, we propose the dataset CCIHP based on the multi-HP dataset CIHP, with 20 new labels covering these 3 kinds of characteristics. In addition, we propose HPTR, a new bottom-up multi-task method based on transformers as a fast and scalable baseline. It is the fastest method of multi-HP state of the art while having precision comparable to the most precise bottom-up method. We hope this will encourage research for fast and accurate methods of precise human descriptions.
