Landmark labeling of training images is essential for many learning tasks in computer vision, such as object detection, tracking, and alignment. Image labeling is typically conduc...
Yan Tong, Xiaoming Liu 0002, Frederick W. Wheeler,...
The goal of semi-supervised learning (SSL) methods is to reduce the amount of labeled training data required by learning from both labeled and unlabeled instances. Macskassy and Pr...
This paper explores the use of the homotopy method for training a semi-supervised Hidden Markov Model (HMM) used for sequence labeling. We provide a novel polynomial-time algorith...
Supervised sequence-labeling systems in natural language processing often suffer from data sparsity because they use word types as features in their prediction tasks. Consequently...
In semi-supervised learning, a number of labeled examples are usually required for training an initial weakly useful predictor which is in turn used for exploiting the unlabeled e...