Faster Image2Video Generation: A Closer Look at CLIP Image Embedding's Impact on Spatio-Temporal Cross-Attentions.
Ashkan TaghipourMorteza GhahremaniMohammed BennamounAref Miri RekavandiZinuo LiHamid LagaFarid BoussaïdPublished in: CoRR (2024)
Keyphrases
- single image
- image content
- image data
- image features
- input image
- image classification
- image retrieval
- spatio temporal
- segmentation method
- low level
- image representation
- multiscale
- region of interest
- image analysis
- image segmentation
- edge detection
- image collections
- high resolution
- image regions
- static images
- feature points
- video data
- image matching
- test images
- temporal domain
- image derivatives