Exploiting Scene Context for Image Captioning.
Rakshith ShettyHamed Rezazadegan TavakoliJorma LaaksonenPublished in: iV&L-MM@MM (2016)
Keyphrases
- single image
- input image
- image regions
- visual context
- complex scenes
- multiscale
- scene images
- image features
- image data
- scene understanding
- image content
- image representation
- scene classification
- image frames
- multiple images
- image segmentation
- imaging process
- scene matching
- camera images
- image retrieval
- visual data
- multiple objects
- spatial relations
- edge detection
- real world scenes
- cast shadows
- reference images
- camera positions
- image based rendering
- pixel values
- image motion
- image classification
- high resolution
- object models
- d scene
- image sequences
- image set
- three dimensional
- image segments
- piecewise planar
- image transformations
- geometric information
- outdoor scenes
- spatial information
- image matching
- vanishing points
- test images
- dynamic scenes