Scenimefy: Learning to Craft Anime Scene via Semi-Supervised Image-to-Image Translation.
Yuxin JiangLiming JiangShuai YangChen Change LoyPublished in: CoRR (2023)
Keyphrases
- single image
- image content
- input image
- multiscale
- image segmentation
- image retrieval
- image data
- image regions
- image set
- test images
- image classification
- image features
- supervised learning
- image frames
- image representation
- low level
- segmentation method
- learning algorithm
- scene geometry
- scene images
- geometric information
- vanishing points
- complex scenes
- image segments
- video sequences
- piecewise planar
- image matching
- spatial information
- image analysis
- active learning
- real world scenes
- scene matching
- scene classification
- multiple images
- pixel values
- spatial relations
- similarity measure
- training set
- feature points