Object Ordering with Bidirectional Matchings for Visual Reasoning

TL;DR

This paper introduces a novel neural model for visual reasoning that uses bidirectional attention and reinforcement learning to accurately match objects with language in complex images, improving performance on the NLVR dataset.

Contribution

The paper presents an end-to-end neural approach combining bidirectional attention and RL-based pointer networks for object ordering in visual reasoning tasks, a novel method for the NLVR dataset.

Findings

Achieves 4-6% absolute improvement over previous state-of-the-art methods.

Effectively matches object orderings with natural language descriptions.

Demonstrates strong performance on both structured and raw image datasets.

Abstract

Visual reasoning with compositional natural language instructions, e.g., based on the newly-released Cornell Natural Language Visual Reasoning (NLVR) dataset, is a challenging task, where the model needs to have the ability to create an accurate mapping between the diverse phrases and the several objects placed in complex arrangements in the image. Further, this mapping needs to be processed to answer the question in the statement given the ordering and relationship of the objects across three similar images. In this paper, we propose a novel end-to-end neural model for the NLVR task, where we first use joint bidirectional attention to build a two-way conditioning between the visual information and the language phrases. Next, we use an RL-based pointer network to sort and process the varying number of unordered objects (so as to match the order of the statement phrases) in each of the three images and then pool over the three decisions. Our model achieves strong improvements (of 4-6% absolute) over the state-of-the-art on both the structured representation and raw image versions of the dataset.