IDAdapter: Learning Mixed Features for Tuning-Free Personalization of Text-to-Image Models.
Siying CuiJia GuoXiang AnJiankang DengYongle ZhaoXinyu WeiZiyong FengPublished in: CoRR (2024)
Keyphrases
- image features
- learned models
- learning algorithm
- low level
- prior knowledge
- learning process
- learning models
- extracted features
- image data
- extracting features
- auto annotation
- image statistics
- image pixels
- multiscale
- feature extraction
- classification models
- keypoints
- image segmentation
- test images
- image content
- single image
- image classification
- segmentation method
- visual object recognition
- image analysis
- feature values
- automatically extracted
- feature representation
- high resolution
- probabilistic model
- input image
- image collections
- edge detection
- spatial information
- natural images
- co occurrence
- invariant features
- web images
- feature set
- global features
- image regions
- feature vectors
- global shape
- keywords
- similarity measure