Abstract
Large amounts of data are widely stored in cyberspace. Not only can they bring much convenience to people’s lives and work, but they can also assist the work in the information security field, such as microexpression recognition and sentiment analysis in the criminal investigation. Thus, it is of great significance to recognize and analyze the sentiment information, which is usually described by different modalities. Due to the correlation among different modalities data, multimodal can provide more comprehensive and robust information than unimodal in data analysis tasks. The complementary information from different modalities can be obtained by multimodal fusion methods. These approaches can process multimodal data through fusion algorithms and ensure the accuracy of the information used for subsequent classification or prediction tasks. In this study, a two-level multimodal fusion (TlMF) method with both data-level and decision-level fusion is proposed to achieve the sentiment analysis task. In the data-level fusion stage, a tensor fusion network is utilized to obtain the text-audio and text-video embeddings by fusing the text with audio and video features, respectively. During the decision-level fusion stage, the soft fusion method is adopted to fuse the classification or prediction results of the upstream classifiers, so that the final classification or prediction results can be as accurate as possible. The proposed method is tested on the CMU-MOSI, CMU-MOSEI, and IEMOCAP datasets, and the empirical results and ablation studies confirm the effectiveness of TlMF in capturing useful information from all the test modalities.
Highlights
Large amounts of data are widely stored in cyberspace
During the decision-level fusion stage, the soft fusion method is adopted to fuse the classification or prediction results of the upstream classifiers, so that the final classification or prediction results can be as accurate as possible. e proposed method is tested on the CMU-MOSI, CMU-MOSEI, and Interactive Emotional Dyadic Motion Capture (IEMOCAP) datasets, and the empirical results and ablation studies confirm the effectiveness of twolevel multimodal fusion (TlMF) in capturing useful information from all the test modalities
Motivated by the above discussion, this study proposes a new modal fusion method named TlMF, which produces unimodal embeddings by using a convolution neural network (CNN)-BiLSTM neural network and achieves the information fusion of text and audio/video embedding based on tensor fusion and decision fusion stages. e cores of our method include (1) a tensor fusion network used to fuse text data with video and audio data, respectively, and (2) a decision-level fusion strategy, which can fuse the classification results. en, three datasets of multimodal sentiment analysis and emotion classification, i.e., CMU-MOSI [32], CMU-MOSEI [33], and IEMOCAP [34] are used for experiments to evaluate the effectiveness of the proposed methods
Summary
Large amounts of data are widely stored in cyberspace. can they bring much convenience to people’s lives and work, but they can assist the work in the information security field, such as microexpression recognition and sentiment analysis in the criminal investigation. us, it is of great significance to recognize and analyze the sentiment information, which is usually described by different modalities. Based on the TFN, low-rank multimodal fusion (LMF) [17], memory fusion network (MFN) [18], and multimodal transformer (MuLT) [19] have been proposed, which can further improve the processing efficiency and evaluation It can be seen from these results that attaching both audio and video features to the same textual information can enable nontext information to be better understood, and in turn, the nontext information can impart greater meaning to the text [20]
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
