Timezone: »

 
Poster
ButterflyFlow: Building Invertible Layers with Butterfly Matrices
Chenlin Meng · Linqi Zhou · Kristy Choi · Tri Dao · Stefano Ermon

Tue Jul 19 03:30 PM -- 05:30 PM (PDT) @ Hall E #322
in Poster Session 1 »

Normalizing flows model complex probability distributions using maps obtained by composing invertible layers. Special linear layers such as masked and 1×1 convolutions play a key role in existing architectures because they increase expressive power while having tractable Jacobians and inverses. We propose a new family of invertible linear layers based on butterfly layers, which are known to theoretically capture complex linear structures including permutations and periodicity, yet can be inverted efficiently. This representational power is a key advantage of our approach, as such structures are common in many real-world datasets. Based on our invertible butterfly layers, we construct a new class of normalizing flow mod- els called ButterflyFlow. Empirically, we demonstrate that ButterflyFlows not only achieve strong density estimation results on natural images such as MNIST, CIFAR-10, and ImageNet-32×32, but also obtain significantly better log-likelihoods on structured datasets such as galaxy images and MIMIC-III patient cohorts—all while being more efficient in terms of memory and computation than relevant baselines.

Keywords: DL: Algorithms · DL: Everything Else · DL: Generative Models and Autoencoders

Author Information

Chenlin Meng (Stanford University)
Linqi Zhou (Stanford University)
Kristy Choi (Stanford University)
Tri Dao (Stanford)
Stefano Ermon (Stanford University)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors