Multi-modality Latent Interaction Network for Visual Question Answering

TL;DR

This paper introduces the Multi-modality Latent Interaction module (MLI) for VQA, which models cross-modality relationships between summarized visual and language features, leading to improved performance on benchmark datasets.

Contribution

The paper proposes a novel MLI module that captures latent cross-modality relations through summarizations, enhancing VQA performance beyond existing relation modeling methods.

Findings

Achieves competitive results on VQA v2.0 and TDIUC datasets.

Stacking MLI modules models complex relations effectively.

Combining with BERT further improves accuracy.

Abstract

Exploiting relationships between visual regions and question words have achieved great success in learning multi-modality features for Visual Question Answering (VQA). However, we argue that existing methods mostly model relations between individual visual regions and words, which are not enough to correctly answer the question. From humans' perspective, answering a visual question requires understanding the summarizations of visual and language information. In this paper, we proposed the Multi-modality Latent Interaction module (MLI) to tackle this problem. The proposed module learns the cross-modality relationships between latent visual and language summarizations, which summarize visual regions and question into a small number of latent representations to avoid modeling uninformative individual region-word relations. The cross-modality information between the latent summarizations are propagated to fuse valuable information from both modalities and are used to update the visual and word features. Such MLI modules can be stacked for several stages to model complex and latent relations between the two modalities and achieves highly competitive performance on public VQA benchmarks, VQA v2.0 and TDIUC . In addition, we show that the performance of our methods could be significantly improved by combining with pre-trained language model BERT.