StyleEDL: Style-Guided High-order Attention Network for Image Emotion Distribution Learning

TL;DR

This paper introduces StyleEDL, a novel style-guided high-order attention network that leverages hierarchical stylistic information and graph convolutional networks to improve image emotion distribution learning, addressing emotion ambiguity and enhancing interpretability.

Contribution

The paper proposes a new network architecture that integrates stylistic-aware representations and high-order attention mechanisms for more accurate emotion distribution learning from images.

Findings

Outperforms state-of-the-art methods on benchmark datasets.

Effectively captures stylistic and visual part interactions.

Demonstrates robustness across different image datasets.

Abstract

Emotion distribution learning has gained increasing attention with the tendency to express emotions through images. As for emotion ambiguity arising from humans' subjectivity, substantial previous methods generally focused on learning appropriate representations from the holistic or significant part of images. However, they rarely consider establishing connections with the stylistic information although it can lead to a better understanding of images. In this paper, we propose a style-guided high-order attention network for image emotion distribution learning termed StyleEDL, which interactively learns stylistic-aware representations of images by exploring the hierarchical stylistic information of visual contents. Specifically, we consider exploring the intra- and inter-layer correlations among GRAM-based stylistic representations, and meanwhile exploit an adversary-constrained high-order attention mechanism to capture potential interactions between subtle visual parts. In addition, we introduce a stylistic graph convolutional network to dynamically generate the content-dependent emotion representations to benefit the final emotion distribution learning. Extensive experiments conducted on several benchmark datasets demonstrate the effectiveness of our proposed StyleEDL compared to state-of-the-art methods. The implementation is released at: https://github.com/liuxianyi/StyleEDL.