A novel approach to video transition detection based on a two-phase classification strategy.
Shijin LiLin LinXiaofang LiPublished in: ICNSC (2008)
Keyphrases
- pattern recognition
- robust detection
- shot detection
- machine learning
- pattern classification
- decision trees
- classification method
- neyman pearson
- face detection and tracking
- activity detection
- feature space
- classification accuracy
- classification scheme
- text classification
- benchmark datasets
- classification algorithm
- detection algorithm
- support vector machine svm
- training samples
- support vector machine
- feature extraction and classification
- false alarms
- automatic classification
- event detection
- video classification
- false positives
- feature selection
- detection method
- multimedia
- video sequences
- support vector
- model selection
- multi class
- video database
- microcalcification clusters
- computer vision
- tv broadcast
- mammogram images
- feature extraction
- shot boundary detection
- active learning
- video segmentation
- neural network
- anomaly detection
- video analysis
- video clips
- video data
- video content
- key frames
- video streams
- video frames
- space time
- class labels