Pick-and-Draw: Training-free Semantic Guidance for Text-to-Image Personalization.
Henglei LvJiayu XiaoLiang LiQingming HuangPublished in: CoRR (2024)
Keyphrases
- single image
- input image
- image data
- high resolution
- semantic information
- multiscale
- image content
- image features
- image classification
- image analysis
- image regions
- low level
- semantic network
- image segmentation
- textual descriptions
- image representation
- segmentation method
- image collections
- semantic space
- image processing
- edge detection
- natural language
- feature points
- image retrieval
- pixel values
- web images
- semantic representation
- text information
- information retrieval
- semantic categories