Optimising the Input Image to Improve Visual Relationship Detection.

Noel Mizzi Adrian Muscat

Published in: CoRR (2019)

Keyphrases

input image
natural images
binary images
gray level
saliency map
input data
segmentation algorithm
detection algorithm
superpixels
object detection
low level
depth map
image structure
principal component analysis
human faces
visual information
denoising
natural scenes
edge pixels