The Importance of Metric Learning for Robotic Vision: Open Set Recognition and Active Learning.
Benjamin J. MeyerTom DrummondPublished in: CoRR (2019)
Keyphrases
- metric learning
- robotic vision
- active learning
- semi supervised
- distance metric learning
- distance metric
- pairwise
- semi supervised learning
- machine learning and pattern recognition
- recognition rate
- dimensionality reduction
- person re identification
- object recognition
- labeled data
- distance function
- multi task
- learning tasks
- unlabeled data
- random sampling
- feature space
- pattern recognition
- machine learning
- learning algorithm
- recognition algorithm
- selective sampling
- transfer learning
- euclidean distance
- feature extraction
- computer vision
- sample size
- unsupervised learning
- supervised learning
- multi class
- multiple instance learning
- support vector
- image processing
- knowledge base
- mahalanobis metric