Describe me if you can! Characterized Instance-level Human Parsing
arXiv:2201.09594 · doi:10.1109/ICIP42928.2021.9506509
Abstract
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.
arXiv comment5 pages
References in corpus (9)
- Fully Convolutional Networks for Semantic Segmentation
- Improving Person Re-identification by Attribute and Identity Learning
- Clothing Co-Parsing by Joint Image Segmentation and Labeling
- Multiple-Human Parsing in the Wild
- Grapy-ML: Graph Pyramid Mutual Learning for Cross-dataset Human Parsing
- Fashion Meets Computer Vision: A Survey
- ModaNet: A Large-Scale Street Fashion Dataset with Polygon Annotations
- Describe me if you can! Characterized Instance-level Human Parsing
- End-to-end Person Search Sequentially Trained on Aggregated Dataset