Mucko: Multi-Layer Cross-Modal Knowledge Reasoning for Fact-based Visual Question Answering

Abstract

<jats:p>Fact-based Visual Question Answering (FVQA) requires external knowledge beyond the visible content to answer questions about an image. This ability is challenging but indispensable to achieve general VQA. One limitation of existing FVQA solutions is that they jointly embed all kinds of information without fine-grained selection, which introduces unexpected noises for reasoning the final answer. How to capture the question-oriented and information-complementary evidence remains a key challenge to solve the problem. In this paper, we depict an image by a multi-modal heterogeneous graph, which contains multiple layers of information corresponding to the visual, semantic and factual features. On top of the multi-layer graph representations, we propose a modality-aware heterogeneous graph convolutional network to capture evidence from different layers that is most relevant to the given question. Specifically, the intra-modal graph convolution selects evidence from each modality and cross-modal graph convolution aggregates relevant information across different graph layers. By stacking this process multiple times, our model performs iterative reasoning across three modalities and predicts the optimal answer by analyzing all question-oriented evidence. We achieve a new state-of-the-art performance on the FVQA task and demonstrate the effectiveness and interpretability of our model with extensive experiments.</jats:p>

Visual question answering with gated relation‐aware auxiliary

“…OK‐VQA [24] provided a new data set including more than 14,000 questions that require external knowledge to answer. Mucko [25] and [26] utilised the graph structure to capture information from the external fact space for reasoning the answer.…”

X-GGM: Graph Generative Modeling for Out-of-distribution Generalization in Visual Question Answering

“…Graphs are non-Euclidean structured data, which can effectively represent relationships between nodes. Some recent works construct graphs for visual or linguistic elements in V+L tasks, such as VQA [16,27,43], VideoQA [28,30,78], Image Captioning [23,69,75], and Visual Grounding [31,47,68], to reveal relationships between these elements and obtain fine-grained semantic representations. These constructed graphs can be broadly grouped into three types: visual graphs between image objects/regions (e.g., [69]), linguistic graphs between sentence elements/tokens (e.g., [33]), and crossmodal graphs among visual and linguistic elements (e.g., [47]).…”

Worldly Wise (WoW) - Cross-Lingual Knowledge Fusion for Fact-based Visual Spoken-Question Answering

“…In (Wang et al, 2018), FVQA is approached as a parsing and factretrieval problem, while directly retrieves facts using lexical-semantic word embeddings. In Out-of-the-box (OOB) reasoning , a Graph Convolutional Network (Kipf and Welling, 2017) is used to reason about the correct entity, while (Zhu et al, 2020) (the current State-of-the-Art in the complete-KG FVQA task) added a visual scenegraph (Krishna et al, 2016) and a semantic graph based on the question alongside the (OOB) KG reasoning module. In (Ramnath and Hasegawa-Johnson, 2020), FVQA is tackled on incomplete KGs using KG embeddings to represent entities instead of word-embeddings, as the latter are shown to be inadequate for this task.…”