Are Multimodal Models Robust to Image and Text Perturbations?
Jielin QiuYi ZhuXingjian ShiFlorian WenzelZhiqiang TangDing ZhaoBo LiMu LiPublished in: CoRR (2022)
Keyphrases
- input image
- image data
- image content
- random fields
- image noise
- single image
- image representation
- image features
- partial occlusion
- image retrieval
- image segmentation
- image classification
- image pixels
- multi modal
- information retrieval
- probabilistic model
- multiscale
- bayesian framework
- region of interest
- low level
- image statistics
- geometric distortions
- high resolution
- keywords
- test images
- template matching
- image collections
- edge detection
- similarity measure
- feature descriptors
- bounding box
- visual effects
- salt pepper
- image matching
- hough transform
- statistical model
- image analysis
- feature points
- denoising