StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learners.
Yonglong TianLijie FanPhillip IsolaHuiwen ChangDilip KrishnanPublished in: NeurIPS (2023)
Keyphrases
- visual representation
- image data
- single image
- input image
- image retrieval
- random fields
- image analysis
- bayesian framework
- image classification
- object parts
- image collections
- image features
- multiscale
- web images
- high resolution
- image segmentation
- image representation
- segmentation method
- image regions
- artificial intelligence
- learning experience
- learning activities
- feature points
- edge detection
- text mining
- e learning
- information retrieval
- learning materials
- language learning
- concept maps
- image processing
- object models
- visual representations