A Transformer-Based Cross-Modal Image-Text Retrieval Method using Feature Decoupling and Reconstruction.
Huan ZhangYingzhi SunYu LiaoSiyuan XuRui YangShuang WangBiao HouLicheng JiaoPublished in: IGARSS (2022)
Keyphrases
- retrieval method
- image retrieval
- cross modal
- content based image
- image features
- image representation
- image data
- image content
- multi modal
- color histogram
- image classification
- multiscale
- image collections
- low level
- visual similarity
- text retrieval
- image segmentation
- image database
- image regions
- keywords
- image search
- high dimensional
- image annotation
- keypoints
- computer vision
- similarity measure
- web images
- multimedia retrieval
- visual features
- feature selection