• Structural Engineering and Mechanics
• /
• v.31 no.3
• /
• pp.333-347
• /
• 2009

Due to structural complicacy, structural health monitoring for civil engineering needs more accurate and effectual methods of damage identification. This study aims to import multi-source information fusion (MSIF) into structural damage diagnosis to improve the validity of damage detection. Firstly, the essential theory and applied mathematic methods of MSIF are introduced. And then, the structural damage identification method based on multi-mode information fusion is put forward. Later, on the basis of a numerical simulation of a concrete continuous box beam bridge, it is obviously indicated that the improved modal strain energy method based on multi-mode information fusion has nicer sensitivity to structural initial damage and favorable robusticity to noise. Compared with the classical modal strain energy method, this damage identification method needs much less modal information to detect structural initial damage. When the noise intensity is less than or equal to 10%, this method can identify structural initial damage well and truly. In a word, this structural damage identification method based on multi-mode information fusion has better effects of structural damage identification and good practicability to actual structures.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.1
• /
• pp.51-73
• /
• 2023

Social media play a significant role in communicating information across the globe, connecting with loved ones, getting the news, communicating ideas, etc. However, a group of people uses social media to spread fake information, which has a bad impact on society. Therefore, minimizing fake news and its detection are the two primary challenges that need to be addressed. This paper presents a multi-modal deep learning technique to address the above challenges. The proposed modal can use and process visual and textual features. Therefore, it has the ability to detect fake information from visual and textual data. We used EfficientNetB0 and a sentence transformer, respectively, for detecting counterfeit images and for textural learning. Feature embedding is performed at individual channels, whilst fusion is done at the last classification layer. The late fusion is applied intentionally to mitigate the noisy data that are generated by multi-modalities. Extensive experiments are conducted, and performance is evaluated against state-of-the-art methods. Three real-world benchmark datasets, such as MediaEval (Twitter), Weibo, and Fakeddit, are used for experimentation. Result reveals that the proposed modal outperformed the state-of-the-art methods and achieved an accuracy of 86.48%, 82.50%, and 88.80%, respectively, for MediaEval (Twitter), Weibo, and Fakeddit datasets.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.12
• /
• pp.505-518
• /
• 2023

3D point cloud semantic segmentation is a computer vision task that involves dividing the point cloud into different objects and regions by predicting the class label of each point. Existing 3D semantic segmentation models have some limitations in performing sufficient fusion of multi-modal features while ensuring both characteristics of 2D visual features extracted from RGB images and 3D geometric features extracted from point cloud. Therefore, in this paper, we propose MMCA-Net, a novel 3D semantic segmentation model using 2D-3D multi-modal features. The proposed model effectively fuses two heterogeneous 2D visual features and 3D geometric features by using an intermediate fusion strategy and a multi-modal cross attention-based fusion operation. Also, the proposed model extracts context-rich 3D geometric features from input point cloud consisting of irregularly distributed points by adopting PTv2 as 3D geometric encoder. In this paper, we conducted both quantitative and qualitative experiments with the benchmark dataset, ScanNetv2 in order to analyze the performance of the proposed model. In terms of the metric mIoU, the proposed model showed a 9.2% performance improvement over the PTv2 model using only 3D geometric features, and a 12.12% performance improvement over the MVPNet model using 2D-3D multi-modal features. As a result, we proved the effectiveness and usefulness of the proposed model.

• Nuclear Engineering and Technology
• /
• v.55 no.10
• /
• pp.3844-3853
• /
• 2023

Although radiation and chemotherapy methods for cancer therapy have advanced significantly, surgical resection is still recommended for most cancers. Therefore, intraoperative imaging studies have emerged as a surgical tool for identifying tumor margins. Intraoperative imaging has been examined using conventional imaging devices, such as optical near-infrared probes, gamma probes, and ultrasound devices. However, each modality has its limitations, such as depth penetration and spatial resolution. To overcome these limitations, hybrid imaging modalities and tracer studies are being developed. In a previous study, a multi-modal laparoscope with silicon photo-multiplier (SiPM)-based gamma detection acquired a 1 s interval gamma image. However, improvements in the near-infrared fluorophore (NIRF) signal intensity and gamma image central defects are needed to further evaluate the usefulness of multi-modal systems. In this study, an attempt was made to change the NIRF image acquisition method and the SiPM-based gamma detector to improve the source detection ability and reduce the image acquisition time. The performance of the multi-modal system using a complementary metal oxide semiconductor and modified SiPM gamma detector was evaluated in a phantom test. In future studies, a multi-modal system will be further optimized for pilot preclinical studies.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.526-527
• /
• 2023

Speech emotion recognition(SER) is one of the interesting topics in the machine learning field. By developing multi-modal speech emotion recognition system, we can get numerous benefits. This paper explain about fusing BERT as the text recognizer and CNN as the speech recognizer to built a multi-modal SER system.

• Annual Conference on Human and Language Technology
• /
• 2022.10a
• /
• pp.95-99
• /
• 2022

오픈 도메인 질의 응답 (ODQA, Open-Domain Question Answering)은 주어진 질문에 대한 답을 찾는 과업으로서 질문과 관련있는 지식을 찾는 "검색" 단계를 필요로 한다. 최근 이미지, 테이블 등의 검색을 요구하는 멀티 모달 ODQA에 대한 연구가 많이 진행되었을 뿐만 아니라 산업에서의 중요도 또한 높아지고 있다. 본 논문은 여러 종류의 멀티 모달 ODQA 중에서도 테이블 - 텍스트 기반 멀티 모달 ODQA 데이터 집합으로 Fusion-in-Decoder (FiD)를 이용한 멀티 모달 오픈 도메인 질의 응답 연구를 제안하며 베이스라인 대비 최대 EM 20.5, F1 23.2 향상을 보였다.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.4
• /
• pp.76-81
• /
• 2009

We propose an effective emotion recognition algorithm based on the minimum classification error (MCE) incorporating multi-modal system The emotion recognition is performed based on a Gaussian mixture model (GMM) based on MCE method employing on log-likelihood. In particular, the reposed technique is based on the fusion of feature vectors based on voice signal and galvanic skin response (GSR) from the body sensor. The experimental results indicate that performance of the proposal approach based on MCE incorporating the multi-modal system outperforms the conventional approach.

• Journal of the Ergonomics Society of Korea
• /
• v.31 no.4
• /
• pp.499-505
• /
• 2012

Objective: Research on the transformation of interaction between mobile machines and users through analysis on current gesture interface technology development trend. Background: For smooth interaction between machines and users, interface technology has evolved from "command line" to "mouse", and now "touch" and "gesture recognition" have been researched and being used. In the future, the technology is destined to evolve into "multi-modal", the fusion of the visual and auditory senses and "3D multi-modal", where three dimensional virtual world and brain waves are being used. Method: Within the development of computer interface, which follows the evolution of mobile machines, actively researching gesture interface and related technologies' trend and development will be studied comprehensively. Through investigation based on gesture based information gathering techniques, they will be separated in four categories: sensor, touch, visual, and multi-modal gesture interfaces. Each category will be researched through technology trend and existing actual examples. Through this methods, the transformation of mobile machine and human interaction will be studied. Conclusion: Gesture based interface technology realizes intelligent communication skill on interaction relation ship between existing static machines and users. Thus, this technology is important element technology that will transform the interaction between a man and a machine more dynamic. Application: The result of this study may help to develop gesture interface design currently in use.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.10
• /
• pp.2768-2787
• /
• 2023

As an important research direction of the application of computer science in the medical field, the automatic generation technology of radiology report has attracted wide attention in the academic community. Because the proportion of normal regions in radiology images is much larger than that of abnormal regions, words describing diseases are often masked by other words, resulting in significant feature loss during the calculation process, which affects the quality of generated reports. In addition, the huge difference between visual features and semantic features causes traditional multi-modal fusion method to fail to generate long narrative structures consisting of multiple sentences, which are required for medical reports. To address these challenges, we propose a multi-feature optimization Transformer (MFOT) for generating radiology reports. In detail, a multi-dimensional mapping attention (MDMA) module is designed to encode the visual grid features from different dimensions to reduce the loss of primary features in the encoding process; a feature pre-fusion (FP) module is constructed to enhance the interaction ability between multi-modal features, so as to generate a reasonably structured radiology report; a detail enhanced attention (DEA) module is proposed to enhance the extraction and utilization of key features and reduce the loss of key features. In conclusion, we evaluate the performance of our proposed model against prevailing mainstream models by utilizing widely-recognized radiology report datasets, namely IU X-Ray and MIMIC-CXR. The experimental outcomes demonstrate that our model achieves SOTA performance on both datasets, compared with the base model, the average improvement of six key indicators is 19.9% and 18.0% respectively. These findings substantiate the efficacy of our model in the domain of automated radiology report generation.

• Journal of Broadcast Engineering
• /
• v.23 no.3
• /
• pp.351-360
• /
• 2018

Human emotion recognition is a research topic that is receiving continuous attention in computer vision and artificial intelligence domains. This paper proposes a method for classifying human emotions through multiple neural networks based on multi-modal signals which consist of image, landmark, and audio in a wild environment. The proposed method has the following features. First, the learning performance of the image-based network is greatly improved by employing both multi-task learning and semi-supervised learning using the spatio-temporal characteristic of videos. Second, a model for converting 1-dimensional (1D) landmark information of face into two-dimensional (2D) images, is newly proposed, and a CNN-LSTM network based on the model is proposed for better emotion recognition. Third, based on an observation that audio signals are often very effective for specific emotions, we propose an audio deep learning mechanism robust to the specific emotions. Finally, so-called emotion adaptive fusion is applied to enable synergy of multiple networks. The proposed network improves emotion classification performance by appropriately integrating existing supervised learning and semi-supervised learning networks. In the fifth attempt on the given test set in the EmotiW2017 challenge, the proposed method achieved a classification accuracy of 57.12%.