Object-Centric Unsupervised Image Captioning.
Zihang MengDavid YangXuefei CaoAshish ShahSer-Nam LimPublished in: CoRR (2021)
Keyphrases
- image analysis
- spatial relations
- input image
- image features
- image regions
- single image
- image data
- region of interest
- image retrieval
- image classification
- image content
- keypoints
- partial occlusion
- feature points
- spatial relationships
- bounding box
- position and orientation
- multiple objects
- similar objects
- normalized correlation
- lighting conditions
- object localization
- multiscale
- edge detection
- segmentation method
- image matching
- image segments
- target object
- image representation
- object shapes
- location and orientation
- spatial information
- test images
- complex scenes
- image segmentation
- surface shape
- pixel level
- high resolution
- face recognition
- object shape
- visual appearance
- background clutter
- relative position
- object models
- semi supervised
- ground plane
- image quality
- object model
- image set