Toward Discriminating and Synthesizing Motion Traces Using Deep Probabilistic Generative Models.
Fan ZhouXin LiuKunpeng ZhangGoce TrajcevskiPublished in: IEEE Trans. Neural Networks Learn. Syst. (2021)
Keyphrases
- generative model
- probabilistic model
- mixture model
- image sequences
- maximum entropy principle
- bayesian framework
- posterior probability
- semi supervised
- em algorithm
- latent dirichlet allocation
- prior knowledge
- object categories
- deep belief networks
- conditional random fields
- motion analysis
- discriminative learning
- topic models
- hidden variables
- expectation maximization
- discriminative models
- deep learning
- hierarchical hidden markov models
- learning process
- object recognition
- bayesian networks