Identifying implementation bugs in machine learning based image classifiers using metamorphic testing.
Anurag DwarakanathManish AhujaSamarth SikandRaghotham M. RaoR. P. Jagadeesh Chandra BoseNeville DubashSanjay PodderPublished in: ISSTA (2018)
Keyphrases
- machine learning
- input image
- image features
- image data
- decision trees
- machine learning methods
- machine learning algorithms
- single image
- image analysis
- image representation
- image segmentation
- image retrieval
- machine learning approaches
- data mining
- supervised classification
- multiscale
- image pixels
- image collections
- image classification
- region of interest
- similarity measure
- image content
- test set
- segmentation method
- image regions
- learning algorithm
- low level
- feature selection
- extracted features
- template matching
- spatial information
- test cases
- training data
- software development
- source code
- image matching
- segmentation algorithm
- test images
- support vector
- image set
- edge detection
- training set
- multi class
- active learning