arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.09666cs.CL
收藏

用于情感识别的多模态端到端稀疏模型

Multimodal End-to-End Sparse Model for Emotion Recognition

Center for Artificial Intelligence Research (CAiRE)(人工智能研究中心 (CAiRE));The Hong Kong University of Science and Technology(香港科技大学)

Wenliang Dai, Samuel Cahyawijaya, Zihan Liu, Pascale Fung

摘要:

现有的多模态情感计算任务,如情感识别,一般采用两阶段的管道,首先提取特征表示为每个单一的模态与手工制作的算法,然后执行端到端的学习与提取的功能。然而,提取的特征是固定的,并且不能在不同的目标任务上进一步微调,并且手动寻找特征提取算法不能很好地推广或扩展到不同的任务,这可能导致次优性能。在本文中,我们开发了一个完全端到端的模型,将这两个阶段连接起来,并对其进行联合优化。此外,我们重新构建了当前的数据集,以实现完全的端到端训练。此外,为了减少端到端模型带来的计算开销,我们引入了一个稀疏的跨模态注意力机制的特征提取。实验结果表明,我们完全端到端的模型显着优于目前的国家的最先进的模型的基础上的两个阶段的管道。此外,通过增加稀疏的跨模态注意力,我们的模型可以保持性能与大约一半的计算在特征提取部分。

英文摘要:

Existing works on multimodal affective computing tasks, such as emotion recognition, generally adopt a two-phase pipeline, first extracting feature representations for each single modality with hand-crafted algorithms and then performing end-to-end learning with the extracted features. However, the extracted features are fixed and cannot be further fine-tuned on different target tasks, and manually finding feature extraction algorithms does not generalize or scale well to different tasks, which can lead to sub-optimal performance. In this paper, we develop a fully end-to-end model that connects the two phases and optimizes them jointly. In addition, we restructure the current datasets to enable the fully end-to-end training. Furthermore, to reduce the computational overhead brought by the end-to-end model, we introduce a sparse cross-modal attention mechanism for the feature extraction. Experimental results show that our fully end-to-end model significantly surpasses the current state-of-the-art models based on the two-phase pipeline. Moreover, by adding the sparse cross-modal attention, our model can maintain performance with around half the computation in the feature extraction part.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑