Progressive Learning and Disentanglement of Hierarchical Representations
This work addresses the challenge of preserving all factors of variation in deep generative models for researchers in representation learning, though it is incremental as it builds on hierarchical and progressive learning concepts.
The paper tackles the problem of learning disentangled hierarchical representations in variational auto-encoders (VAEs) by proposing a progressive learning strategy that grows the network architecture from high- to low-level abstractions. It demonstrates improved disentanglement on two benchmark datasets using three metrics, including a new one, with quantitative gains over existing works.
Learning rich representation from data is an important task for deep generative models such as variational auto-encoder (VAE). However, by extracting high-level abstractions in the bottom-up inference process, the goal of preserving all factors of variations for top-down generation is compromised. Motivated by the concept of "starting small", we present a strategy to progressively learn independent hierarchical representations from high- to low-levels of abstractions. The model starts with learning the most abstract representation, and then progressively grow the network architecture to introduce new representations at different levels of abstraction. We quantitatively demonstrate the ability of the presented model to improve disentanglement in comparison to existing works on two benchmark data sets using three disentanglement metrics, including a new metric we proposed to complement the previously-presented metric of mutual information gap. We further present both qualitative and quantitative evidence on how the progression of learning improves disentangling of hierarchical representations. By drawing on the respective advantage of hierarchical representation learning and progressive learning, this is to our knowledge the first attempt to improve disentanglement by progressively growing the capacity of VAE to learn hierarchical representations.