Learning Granularity-Unified Representations for Text-to-Image Person Re-identification.
Zhiyin ShaoXinyu ZhangMeng FangZhifeng LinJian WangChangxing DingPublished in: ACM Multimedia (2022)
Keyphrases
- image features
- input image
- learning algorithm
- image content
- image segmentation
- person re identification
- learning tasks
- image classification
- image retrieval
- supervised learning
- multiscale
- visual recognition
- similarity measure
- active learning
- multi view
- learning problems
- web images
- high resolution
- feature points
- image representation
- object recognition