Bridging the Gap between Object and Image-level Representations for Open-Vocabulary Detection.
Hanoona Abdul RasheedMuhammad MaazMuhammad Uzair KhattakSalman H. KhanFahad Shahbaz KhanPublished in: CoRR (2022)
Keyphrases
- pixel level
- bounding box
- image regions
- image features
- single image
- lighting conditions
- input image
- target object
- spatial relationships
- image analysis
- image content
- higher level
- multiscale
- normalized correlation
- complex scenes
- object representation
- image segmentation
- image data
- object shapes
- object detectors
- multiple objects
- image matching
- object models
- image retrieval
- keypoints
- partial occlusion
- d objects
- background clutter
- color histogram
- object level
- region of interest
- edge detection
- image classification
- cluttered scenes
- cluttered background
- similar objects
- detecting objects
- moving objects
- low level
- spatial relations
- image set
- object detection
- spatial information
- segmentation method
- color images
- high resolution
- object model
- image collections
- image segments
- detection algorithm
- visual features
- segmentation algorithm