ACORT: A Compact Object Relation Transformer for Parameter Efficient Image Captioning.
Jia Huei TanYing Hua TanChee Seng ChanJoon Huang ChuahPublished in: CoRR (2022)
Keyphrases
- keypoints
- single image
- image analysis
- lighting conditions
- image data
- input image
- image features
- image regions
- image content
- image retrieval
- multiple objects
- bounding box
- region of interest
- image matching
- test images
- multiscale
- spatial relationships
- segmentation method
- image classification
- edge detection
- position and orientation
- background clutter
- visual appearance
- location and orientation
- normalized correlation
- image representation
- low level
- high resolution
- target object
- feature points
- d objects
- similar objects
- object features
- image collections
- color histogram
- complex scenes
- object models
- image segments
- object shapes
- boundary contour
- moving objects