ReFACT: Updating Text-to-Image Models by Editing the Text Encoder.
Dana AradHadas OrgadYonatan BelinkovPublished in: NAACL-HLT (2024)
Keyphrases
- image data
- image features
- input image
- information retrieval
- web images
- text information
- image segmentation
- multiscale
- text mining
- image classification
- textual information
- image representation
- scanned documents
- image editing
- text retrieval
- text documents
- hough transform
- single image
- image analysis
- image retrieval
- keywords
- multiresolution
- complex background
- textual descriptions
- textual and visual information