CLIP based visual-language saliency heatmaps
The idea is rather simple: aggregated cosine similarities of text embedding and image embeddings from random crops.
-
Notifications
You must be signed in to change notification settings - Fork 0
razzant/VLSaliency
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
CLIP based visual-language saliency heatmaps
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published