Content-Style Decoupling for Unsupervised Makeup Transfer without Generating Pseudo Ground Truth.
Zhaoyang SunShengwu XiongYaxiong ChenYi RongPublished in: CoRR (2024)
Keyphrases
- user experience
- ground truth
- input output
- high quality
- generation process
- unsupervised learning
- gold standard
- data sets
- genetic algorithm
- quantitative evaluation
- writing style
- automatically generating
- unsupervised manner
- segmented images
- digital content
- knowledge transfer
- test images
- transfer learning
- semantic information
- supervised learning
- active learning
- digital libraries
- case study
- multimedia
- neural network