Robust Ordinal VAE: Employing Noisy Pairwise Comparisons for Disentanglement
This addresses the challenge of noisy data in disentanglement tasks, particularly for domain-specific applications like medical severity assessment, though it is incremental as it builds on prior VAE work.
The paper tackles the problem of disentangling factors of interest in Variational Autoencoders by using noisy pairwise ordinal comparisons as an inductive bias, and shows that their Robust Ordinal VAE method outperforms existing approaches and is more robust to noise in both benchmarks and a real-world application.
Recent work by Locatello et al. (2018) has shown that an inductive bias is required to disentangle factors of interest in Variational Autoencoder (VAE). Motivated by a real-world problem, we propose a setting where such bias is introduced by providing pairwise ordinal comparisons between instances, based on the desired factor to be disentangled. For example, a doctor compares pairs of patients based on the level of severity of their illnesses, and the desired factor is a quantitive level of the disease severity. In a real-world application, the pairwise comparisons are usually noisy. Our method, Robust Ordinal VAE (ROVAE), incorporates the noisy pairwise ordinal comparisons in the disentanglement task. We introduce non-negative random variables in ROVAE, such that it can automatically determine whether each pairwise ordinal comparison is trustworthy and ignore the noisy comparisons. Experimental results demonstrate that ROVAE outperforms existing methods and is more robust to noisy pairwise comparisons in both benchmark datasets and a real-world application.