MaskCLIP: Masked Self-Distillation Advances Contrastive Language-Image Pretraining.
Xiaoyi DongJianmin BaoYinglin ZhengTing ZhangDongdong ChenHao YangMing ZengWeiming ZhangLu YuanDong ChenFang WenNenghai YuPublished in: CVPR (2023)
Keyphrases
- input image
- image data
- image analysis
- single image
- image features
- image content
- multiscale
- image representation
- image retrieval
- image classification
- test images
- template matching
- post processing
- image collections
- similarity measure
- image pixels
- segmentation method
- high resolution
- pixel level
- lighting conditions
- region of interest
- image matching
- image regions
- keypoints
- feature points
- language learning
- image segmentation
- aerial images
- artificial intelligence