Myvideo

Guest

Login

Composing graphical models with neural networks

Uploaded By: Myvideo
94 views
0
0 votes
0

How to combine the complementary strengths of probabilistic graphical models and neural networks? We compose latent graphical models with neural network observation likelihoods. For inference, we use recognition networks to produce local evidence potentials, then combine them using efficient message-passing algorithms. All components are trained simultaneously with a single stochastic variational inference objective. We use this framework to automatically segment and categorize mouse behavior from raw depth

Share with your friends

Link:

Embed:

Video Size:

Custom size:

x

Add to Playlist:

Favorites
My Playlist
Watch Later