Utilizing Every Image Object for Semi-supervised Phrase Grounding.
Haidong ZhuArka SadhuZhaoheng ZhengRam NevatiaPublished in: WACV (2021)
Keyphrases
- semi supervised
- bounding box
- image data
- image segmentation
- region of interest
- image retrieval
- image content
- single image
- complex scenes
- image classification
- image regions
- spatial relationships
- target object
- image analysis
- visual appearance
- spatial relations
- normalized correlation
- image representation
- input image
- image features
- keypoints
- location and orientation
- object features
- segmentation algorithm
- test images
- foreground and background
- position and orientation
- multiscale
- multiple objects
- high resolution
- segmentation method
- partial occlusion
- image matching
- object segmentation
- pixel level
- object models
- labeled data
- similarity measure
- back projection
- similar objects
- image set
- image segments
- computer vision