Structured Output Learning with Conditional Generative Flows

TL;DR

This paper introduces c-Glow, a conditional generative flow model that efficiently learns structured output distributions without surrogate objectives, enabling exact likelihood computation and effective sampling for diverse structured prediction tasks.

Contribution

The paper proposes c-Glow, a novel flow-based model for structured prediction that allows exact likelihood computation and simplifies training and inference processes.

Findings

Outperforms state-of-the-art baselines on several tasks

Allows efficient and exact likelihood computation

Demonstrates versatility across multiple structured prediction problems

Abstract

Traditional structured prediction models try to learn the conditional likelihood, i.e., p(y|x), to capture the relationship between the structured output y and the input features x. For many models, computing the likelihood is intractable. These models are therefore hard to train, requiring the use of surrogate objectives or variational inference to approximate likelihood. In this paper, we propose conditional Glow (c-Glow), a conditional generative flow for structured output learning. C-Glow benefits from the ability of flow-based models to compute p(y|x) exactly and efficiently. Learning with c-Glow does not require a surrogate objective or performing inference during training. Once trained, we can directly and efficiently generate conditional samples. We develop a sample-based prediction method, which can use this advantage to do efficient and effective inference. In our experiments, we test c-Glow on five different tasks. C-Glow outperforms the state-of-the-art baselines in some tasks and predicts comparable outputs in the other tasks. The results show that c-Glow is versatile and is applicable to many different structured prediction problems.