Cascaded Mutual Modulation for Visual Reasoning

TL;DR

The paper introduces CMM, a novel multi-step visual reasoning model that uses mutual modulation between text and image features, achieving state-of-the-art results on CLEVR and NLVR benchmarks.

Contribution

CMM is the first to incorporate cascaded mutual modulation with multi-step reasoning for improved visual question answering.

Findings

CMM outperforms previous models on CLEVR and NLVR benchmarks.

Multi-step reasoning and visual-guided language modulation are crucial for performance.

Ablation studies validate the effectiveness of the proposed framework.

Abstract

Visual reasoning is a special visual question answering problem that is multi-step and compositional by nature, and also requires intensive text-vision interactions. We propose CMM: Cascaded Mutual Modulation as a novel end-to-end visual reasoning model. CMM includes a multi-step comprehension process for both question and image. In each step, we use a Feature-wise Linear Modulation (FiLM) technique to enable textual/visual pipeline to mutually control each other. Experiments show that CMM significantly outperforms most related models, and reach state-of-the-arts on two visual reasoning benchmarks: CLEVR and NLVR, collected from both synthetic and natural languages. Ablation studies confirm that both our multistep framework and our visual-guided language modulation are critical to the task. Our code is available at https://github.com/FlamingHorizon/CMM-VR.