sunpia / deepHMM

A PyTorch implementation of a Deep Hidden Markov Model [Structured Inference Networks for Nonlinear State Space Models]

DHMM

A PyTorch implementation of a Deep Hidden Markov Model: Structured Inference Networks for Nonlinear State Space Models.

Adopted from https://github.com/uber/pyro/tree/dev/examples/dmm
and https://github.com/clinicalml/structuredinference

Dependency

• observations (for dataset)

Dataset

Synthetic

https://github.com/clinicalml/theanomodels/blob/master/datasets/synthp.py

Polyphonic

Related Work

• Structured Inference Networks for Nonlinear State Space Models
• 2019ICLR - A NOVEL VARIATIONAL FAMILY FOR HIDDEN NON-LINEAR MARKOV MODELS
• A Recurrent Latent Variable Model for Sequential Data
• STATE SPACE LSTM MODELS WITH PARTICLE MCMC INFERENCE
• BLACK BOX VARIATIONAL INFERENCE FOR STATE SPACE MODEL
• Gaussian variational approximation for high-dimensional state space models
• Generating Long-term Trajectories Using Deep Hierarchical Networks
• Disentangled Sequential Autoencoder