MuRAL: Multi-Scale Region-based Active Learning for Object Detection.
Yi-Syuan LiouTsung-Han WuJia-Fong YehWen-Chin ChenWinston H. HsuPublished in: CoRR (2023)
Keyphrases
- object detection
- active learning
- multiscale
- image segmentation
- coarse to fine
- face detection
- computer vision
- object class
- selective sampling
- background subtraction
- multi class
- object categories
- learning strategies
- edge detection
- scale space
- training examples
- scene understanding
- pedestrian detection
- object recognition
- learning algorithm
- experimental design
- random sampling
- learning process
- multiple scales
- deep structure
- imbalanced data classification
- object classification
- optic flow
- object segmentation
- human detection
- region matching
- machine learning
- training set
- natural images
- local binary pattern
- pool based active learning
- supervised learning
- semi supervised
- wavelet transform
- active learning strategies
- labeled data
- superpixels
- object detectors
- image fusion
- image retrieval
- batch mode
- image processing
- feature selection
- image representation
- active learner
- semi supervised learning
- data sets