VieCap4H-VLSP 2021: ObjectAoA - Enhancing performance of Object Relation Transformer with Attention on Attention for Vietnamese image captioning.
Nghia Hieu NguyenDuong T. D. VoMinh-Quan HaPublished in: CoRR (2022)
Keyphrases
- attention mechanism
- image data
- image features
- multiscale
- keypoints
- input image
- visual attention
- image content
- image analysis
- single image
- image representation
- computer vision
- image regions
- d objects
- image retrieval
- image segmentation
- bounding box
- target object
- lighting conditions
- image matching
- fuzzy logic
- object recognition
- segmentation method
- pixel level
- position and orientation
- spatial relationships
- image segments
- object model
- complex scenes
- low level
- image classification
- focus of attention
- visual appearance
- location and orientation
- normalized correlation
- object shapes
- relative position
- multiple objects
- partial occlusion
- spatial relations
- spatial information
- edge detection
- image processing