OR-Net: Pointwise Relational Inference for Data Completion under Partial Observation

TL;DR

This paper introduces OR-Net, a relational inference model that effectively completes incomplete data by modeling pointwise relationships within and across observed and unobserved data points, applicable across multiple data modalities.

Contribution

The paper proposes OR-Net, a novel relational inference framework that captures pointwise data relativity for improved data completion under partial observations.

Findings

Effective data completion across various modalities

Generalizes well to different scenarios with partial data

Outperforms existing methods in function, image, and motion data tasks

Abstract

Contemporary data-driven methods are typically fed with full supervision on large-scale datasets which limits their applicability. However, in the actual systems with limitations such as measurement error and data acquisition problems, people usually obtain incomplete data. Although data completion has attracted wide attention, the underlying data pattern and relativity are still under-developed. Currently, the family of latent variable models allows learning deep latent variables over observed variables by fitting the marginal distribution. As far as we know, current methods fail to perceive the data relativity under partial observation. Aiming at modeling incomplete data, this work uses relational inference to fill in the incomplete data. Specifically, we expect to approximate the real joint distribution over the partial observation and latent variables, thus infer the unseen targets respectively. To this end, we propose Omni-Relational Network (OR-Net) to model the pointwise relativity in two aspects: (i) On one hand, the inner relationship is built among the context points in the partial observation; (ii) On the other hand, the unseen targets are inferred by learning the cross-relationship with the observed data points. It is further discovered that the proposed method can be generalized to different scenarios regardless of whether the physical structure can be observed or not. It is demonstrated that the proposed OR-Net can be well generalized for data completion tasks of various modalities, including function regression, image completion on MNIST and CelebA datasets, and also sequential motion generation conditioned on the observed poses.