Detection of Similarity Between Images Based on Contrastive Language-Image Pre-Training Neural Network.
Vasyl LytvynRoman PeleshchakIhor RishnyakBohdan KopachYuriy GalPublished in: COLINS (1) (2024)
Keyphrases
- input image
- image data
- image features
- image collections
- test images
- image retrieval
- image classification
- image similarity
- neural network
- original images
- image analysis
- image regions
- similarity measure
- labeled images
- image content
- region of interest
- object detectors
- image set
- spot detection
- image pixels
- similarity metric
- visual appearance
- set of training images
- pixel values
- bounding box
- edge detection
- segmentation method
- web images
- sample images
- image structure
- circular hough transform
- image database
- intensity variations
- lighting conditions
- visual similarity
- feature points
- hausdorff distance
- image matching
- segmentation algorithm
- synthesized images
- pavement distress
- texture and shape features
- object detection
- aerial images
- complex background
- gray value
- imaging process
- image annotation
- partial occlusion
- keypoints
- semantic labels
- single image
- detecting objects
- training set
- road signs
- visual features
- contrast enhancement
- gaussian noise
- segmented images
- image segmentation
- hough transform
- relevant images
- object categories
- color histogram
- visual concepts