Explanatory instructions, detailed text descriptions of image-to-image transformations, are introduced with a 12M-pair dataset and show qualitative evidence of zero-shot generalization on unseen vision tasks.
For various vision tasks, even when dealing with vision tasks related to a single image pair, different people may interpret and describe the images and tasks differently
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Explanatory Instructions: Towards Unified Vision Tasks Understanding and Zero-shot Generalization
Explanatory instructions, detailed text descriptions of image-to-image transformations, are introduced with a 12M-pair dataset and show qualitative evidence of zero-shot generalization on unseen vision tasks.