BIGbench: A Unified Benchmark for Social Bias in Text-to-Image Generative Models Based on Multi-modal LLM.
Hanjun LuoHaoyu HuangZiye DengXuecheng LiuRuizhe ChenZuozhu LiuPublished in: CoRR (2024)
Keyphrases
- multi modal
- generative model
- multiple modalities
- input image
- image features
- image retrieval
- image content
- uni modal
- discriminative learning
- probabilistic model
- image annotation
- multiscale
- image set
- image classification
- image representation
- single modality
- prior knowledge
- image collections
- test images
- web images
- object parts
- text mining
- image segmentation
- image regions
- cross modal
- image matching
- semi supervised
- low level
- high resolution
- hierarchical hidden markov models
- conditional random fields
- topic models
- segmentation algorithm
- higher level
- active learning
- high dimensional
- similarity measure