Timezone: »

 
Poster
Constraining Variational Inference with Geometric Jensen-Shannon Divergence
Jacob Deasy · Nikola Simidjievski · Pietro Lió

Thu Dec 10 09:00 PM -- 11:00 PM (PST) @ Poster Session 6 #1755
in Poster Session 7 »
We examine the problem of controlling divergences for latent space regularisation in variational autoencoders. Specifically, when aiming to reconstruct example $x\in\mathbb{R}^{m}$ via latent space $z\in\mathbb{R}^{n}$ ($n\leq m$), while balancing this against the need for generalisable latent representations. We present a regularisation mechanism based on the {\em skew-geometric Jensen-Shannon divergence} $\left(\textrm{JS}^{\textrm{G}_{\alpha}}\right)$. We find a variation in $\textrm{JS}^{\textrm{G}_{\alpha}}$, motivated by limiting cases, which leads to an intuitive interpolation between forward and reverse KL in the space of both distributions and divergences. We motivate its potential benefits for VAEs through low-dimensional examples, before presenting quantitative and qualitative results. Our experiments demonstrate that skewing our variant of $\textrm{JS}^{\textrm{G}_{\alpha}}$, in the context of $\textrm{JS}^{\textrm{G}_{\alpha}}$-VAEs, leads to better reconstruction and generation when compared to several baseline VAEs. Our approach is entirely unsupervised and utilises only one hyperparameter which can be easily interpreted in latent space.
Keywords: Online Learning · Control Theory · Algorithms · Theory

Author Information

Jacob Deasy (University of Cambridge)
Nikola Simidjievski (University of Cambridge)
Pietro Lió (University of Cambridge)

More from the Same Authors