TeTIm-Eval: a novel curated evaluation data set for comparing text-to-image models.
Federico A. GalatoloMario G. C. A. CiminoEdoardo CogottiPublished in: CoRR (2022)
Keyphrases
- data sets
- image data
- input image
- image content
- image features
- multiscale
- random fields
- image retrieval
- image pixels
- template matching
- statistical model
- image segmentation
- bayesian framework
- probabilistic model
- edge detection
- image classification
- image representation
- information retrieval
- spatial information
- hough transform
- test images
- low level
- high resolution
- parametric models
- textual information
- web images
- image regions
- text retrieval
- image collections
- pixel values
- image analysis
- bounding box
- text information
- set of training images