Utilizing Every Image Object for Semi-supervised Phrase Grounding.
Haidong ZhuArka SadhuZhaoheng ZhengRam NevatiaPublished in: CoRR (2020)
Keyphrases
- semi supervised
- image data
- image features
- multiscale
- image regions
- image retrieval
- input image
- image content
- single image
- image segmentation
- region of interest
- keypoints
- complex scenes
- image representation
- feature points
- image analysis
- location and orientation
- spatial relationships
- partial occlusion
- lighting conditions
- visual appearance
- image classification
- target object
- edge detection
- moving objects
- multiple objects
- image matching
- pairwise
- pixel level
- object models
- position and orientation
- normalized correlation
- spatial relations
- supervised learning
- object segmentation
- spatial information
- segmentation method
- bounding box
- unlabeled data
- semi supervised learning
- labeled data
- intensity images
- multi view
- similar objects
- d objects
- similarity measure