Discriminative HMMS, log-linear models, and CRFS: What is the difference?
Georg HeigoldSimon WieslerMarkus Nußbaum-ThomPatrick LehnenRalf SchlüterHermann NeyPublished in: ICASSP (2010)
Keyphrases
- log linear models
- conditional random fields
- hidden markov models
- discriminative training
- log linear
- max margin
- crf model
- speech recognition
- information extraction
- mixture model
- higher order
- probabilistic model
- graphical models
- generative model
- markov random field
- structured prediction
- markov networks
- parameter learning
- pairwise
- approximate inference
- image segmentation
- object recognition