Entropy-based active learning for object recognition.
Alex HolubPietro PeronaMichael C. BurlPublished in: CVPR Workshops (2008)
Keyphrases
- active learning
- object recognition
- computer vision
- natural images
- d objects
- random sampling
- learning algorithm
- semi supervised
- relevance feedback
- experimental design
- machine learning
- labeled data
- image understanding
- training examples
- scene classification
- object categorization
- active learning strategies
- visual recognition
- annotation effort
- learning process
- keypoints
- visual object recognition
- selective sampling
- object detection
- object classification
- image representation
- object categories
- template matching
- learning strategies
- feature detection
- information theory
- semi supervised learning
- viewpoint
- batch mode
- biologically motivated
- constrained search
- training set
- object class
- supervised learning
- shape recognition
- multiscale