FAIntbench: A Holistic and Precise Benchmark for Bias Evaluation in Text-to-Image Models.
Hanjun LuoZiye DengRuizhe ChenZuozhu LiuPublished in: CoRR (2024)
Keyphrases
- image classification
- multiscale
- image content
- image analysis
- image data
- single image
- input image
- image representation
- image retrieval
- image features
- bayesian framework
- probabilistic model
- template matching
- image segmentation
- image regions
- statistical model
- text mining
- edge detection
- high resolution
- image pixels
- low level
- text retrieval
- region of interest
- image statistics
- scanned documents
- random fields
- text classification
- parametric models
- textual information
- visual effects
- object models
- textual and visual information
- berkeley segmentation dataset
- information retrieval
- bounding box
- pixel values
- test images
- spatial information
- hough transform
- feature points
- image processing
- computer vision