Comparing the Max and Noisy-Or Pooling Functions in Multiple Instance Learning for Weakly Supervised Sequence Learning Tasks.
Yun WangJuncheng LiFlorian MetzePublished in: INTERSPEECH (2018)
Keyphrases
- learning tasks
- multiple instance learning
- weakly supervised learning
- weakly supervised
- supervised learning
- semi supervised
- learning problems
- machine learning
- multi class
- multi task
- semi supervised learning
- learning algorithm
- transfer learning
- active learning
- object class
- unsupervised learning
- learning experience
- machine learning algorithms
- superpixels
- kernel methods
- topic models
- relation extraction
- multi label
- metric learning
- training data
- training set
- image annotation
- training examples
- labeled data
- pairwise
- computer vision
- data mining
- image segmentation
- class labels
- training samples
- unlabeled data
- probabilistic model