A Simple and Effective Use of Object-Centric Images for Long-Tailed Object Detection.
Cheng ZhangTai-Yu PanYandong LiHexiang HuDong XuanSoravit ChangpinyoBoqing GongWei-Lun ChaoPublished in: CoRR (2021)
Keyphrases
- object detection
- image data
- detecting objects
- target object
- input image
- multiple objects
- lighting conditions
- deformable objects
- rigid body
- objects represented
- image database
- category level
- image regions
- object recognition
- visual appearance
- image features
- acquired images
- object localization
- image classification
- partial occlusion
- object class
- d objects
- real objects
- geometric information
- background clutter
- object detectors
- normalized correlation
- three dimensional objects
- bounding box
- three dimensional
- moving objects
- gaussian distribution
- illumination conditions
- image analysis
- image set
- region of interest
- image retrieval
- image registration
- individual objects
- surface shape
- segmentation algorithm
- keypoints
- multiple images
- object models
- spatial relationships
- pixel level
- complex scenes
- object segmentation
- background subtraction
- test images
- object tracking
- object parts
- spatial relations
- viewing angle
- scene understanding
- feature points
- intensity images