Towards a Unified Foundation Model: Jointly Pre-Training Transformers on Unpaired Images and Text.
Qing LiBoqing GongYin CuiDan KondratyukXianzhi DuMing-Hsuan YangMatthew BrownPublished in: CoRR (2021)
Keyphrases
- three dimensional
- image data
- image features
- image database
- image analysis
- statistical model
- input image
- text detection
- unified model
- geometric information
- image classification
- image registration
- probability distribution
- similarity measure
- high level
- image segmentation
- segmentation algorithm
- image collections
- geometric constraints
- probabilistic model
- observed scene