Papers - Noboru Babaguchi

Cross-Modal Guided Visual Representation Learning for Social Image Retrieval Reviewed International coauthorship

1. Ziyu Guan, Wanqing Zhao, Hongmin Liu, Yuta Nakashima, Noboru Babaguchi, Xiaofei He

IEEE Trans Pattern Analysis and Machine Intelligence   47   2186 - 2198   2025.3

Paladin: Understanding Video Intentions in Political Advertisement Videos Reviewed

1. Hong Liu, Yuta Nakashima, Noboru Babaguchi

Proc. IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)   2025.2

Enhancing Fake News Detection in Social Media via Label Propagation on Cross-modal Tweet Graph Reviewed International coauthorship International journal

Wanqing Zhao,Yuta Nakashima, Haiyuan Chen, Noboru Babaguchi

Proceedings of the 31st ACM International Conference on Multimedia   2400 - 2408   2023.10

　More details

Fake news detection in social media has become increasingly important due to the rapid proliferation of personal media channels and the consequential dissemination of misleading information. Our method constructs a cross-modal tweet graph using CLIP, which encodes images and text into a unified space, allowing us to extract potential connections based on similarities in text and images. We then design a Feature Contextualization Network with Label Propagation (FCN-LP) to model the interaction among tweets as well as positive or negative correlations between predicted labels of connected tweets. The propagated labels from the graph are weighted and aggregated for the final detection. To enhance the model’s generalization ability to unseen events, we
introduce a domain generalization loss that ensures consistent features between tweets on seen and unseen events. We use three publicly available fake news datasets, Twitter, PHEME, and Weibo, for evaluation. Our method consistently improves the performance
over the state-of-the-art methods on all benchmark datasets and effectively demonstrates its aptitude for generalizing fake news detection in social media.

An Experimental Consideration on Gait Spoofing Reviewed

Y. Hirose, K. Nakamura, N. Nitta, and N. Babaguchi

Proc International Conference on Computer Vision Theory and Applications (VISAPP2023)   (8)   2023.2

Social IoT Approach to Cyber Defense of a Deep-Learning-Based Recognition System in Front of Media Clones Generated by Model Inversion Attack Reviewed International coauthorship

Mahdi Khosravy, Kazuaki Nakamura, Naoko Nitta, Nilanjan Dey, Rubén González Crespo, Enrique Herrera-Viedma, and Noboru Babaguchi

IEEE Trans Systems, Man, and Cybernetics: Systems   (11)   2022.11

Anonymization of Human Gait in Video Based on Silhouette Deformation and Texture Transfer Reviewed International journal

Yuki Hirose, Kazuaki Nakamura, Naoko Nitta, and Noboru Babaguchi

IEEE Trans. Information Forensics and Security   17   3375 - 3390   2022.9

　More details

This paper proposes a method for anonymizing the appearance of walking people, namely human gait, in video. In the proposed method, we first crop human regions from all frames in an input video and binarize them to get their silhouettes. Next, we slightly deform the silhouettes from the aspects of static body shape and dynamic walking rhythm so that the person in the input video cannot be correctly identified by gait recognition techniques. After that, the textures of the original human regions are transferred onto the deformed silhouettes. Finally, the anonymized human regions with the transferred textures are filled back into the input video. In the results of our experiments, we successfully degraded the accuracy of CNN-based gait recognition systems from 100% to 1.57% in the lowest case without yielding serious distortion in the appearance of the human regions, which demonstrated the effectiveness of the proposed method.

Effective De-identification Generative Adversarial Network for Face Anonymization Reviewed

Z. Kuang, H. Liu, J. Yu, A. Tian, L. Wang, J. Fan, N. Babaguchi

Proc. of 29th ACM International Conference on Multimedia   3182 - 3192   2021.10

Unnoticeable synthetic face replacement for image privacy protection Reviewed

Z. Kuang, Z. Guo, J. Fang, J. Yu, N. Babaguchi, J. Fan

Neurocomputing   457   322 - 333   2021.10

Model Inversion Attack: Analysis under Gray-box Scenario on Deep Learning based Face Recognition System Reviewed

M. Khosravy, K. Nakamura, Y. Hirose, N. Nitta, N. Babaguchi

KSII Transactions on Internet and Information Systems   15 ( 3 )   1100 - 1118   2021.3

Generation and Detection of Media Clones Invited Reviewed

I. Echizen, N. Babaguchi, J. Yamagishi, N. Nitta, Y. Nakashima, K. Nakamura, K. Kono, F. Fang, S. Myojin, Z. Kuang, H. H. Nguyen, N-D. T. Tieu

IEICE Transactions on Information and Systems   E104-D ( 1 )   12 - 23   2021.1

Preventing Fake Information Generation Against Media Clone Attacks Invited Reviewed

N. Babaguchi, I. Echizen, J. Yamagishi, N. Nitta, Y. Nakashima, K. Nakamura, K. Kono, F. Fang, S. Myojin, Z. Kuang, H. H. Nguyen, N-D. T. Tieu

IEICE Transactions on Information and Systems   E104-D ( 1 )   2 - 11   2021.1

Semi-Supervised Outdoor Image Generation Conditioned on Weather Signals Reviewed

S. Kawakami, K. Okada, N. Nitta, K. Nakamura, N. Babaguchi

Proc. of International Conference on Pattern Recognition (ICPR2021)   4268 - 4275   2021.1

Deep Face Recognizer Privacy Attack: Model Inversion Initialization by a Deep Generative Adversarial Data Space Discriminator Reviewed

M. Khosravy, K. Nakamura, N. Nitta, N. Babaguchi

Proc. of a-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC 2020)   1400 - 1405   2020.12

Detection of Cloned Recognizers: A Defending Method against Recognizer Cloning Attack Reviewed

Y. Mori, K. Nakamura, N. Nitta, N. Babaguchi

Proc. of a-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC 2020)   1375 - 1380   2020.12

Update of Graduate School of Engineering Invited

Noboru Babaguchi

72 ( 3 )   1 - 2   2020.8

Constructing Geospatial Concept Graphs from Tagged Images for Geo-Aware Fine-Grained Image Recognition Reviewed

N. Nitta, K. Nakamura, N. Babaguchi

ISPRS International Journal of Geo-Information   9 ( 6 )   Article No. 354 - 23 pages   2020.5

　More details

While visual appearances play a main role in recognizing the concepts captured in images, additional information can provide complementary information for fine-grained image recognition, where concepts with similar visual appearances such as species of birds need to be distinguished. Especially for recognizing geospatial concepts, which are observed only at specific places, geographical locations of the images can improve the recognition accuracy. However, such geo-aware fine-grained image recognition requires prior information about the visual and geospatial features of each concept or the training data composed of high-quality images for each concept associated with correct geographical locations. By using a large number of images photographed in various places and described with textual tags which can be collected from image sharing services such as Flickr, this paper proposes a method for constructing a geospatial concept graph which contains the necessary prior information for realizing the geo-aware fine-grained image recognition, such as a set of visually recognizable fine-grained geospatial concepts, their visual and geospatial features, and the coarse-grained representative visual concepts whose visual features can be transferred to several fine-grained geospatial concepts. Leveraging the information from the images captured by many people can automatically extract diverse types of geospatial concepts with proper features for realizing efficient and effective geo-aware fine-grained image recognition.

Probabilistic Stone’s Blind Source Separation with application to channel estimation and multi-node identification in MIMO IoT green communication and multimedia systems Reviewed

M. Khosravy, N. Gupta, N. Patel, N. Dey, N. Nitta, N. Babaguchi

Computer Communications, Elsevier   157   423 - 433   2020.5

A logical consideration on fraudulent email communication Reviewed

S. Myojin, N. Babaguchi

ARTIFICIAL LIFE AND ROBOTICS   25 ( 3 )   7pages   2020.3

　More details

One of the most serious problems in modern society is that internet users are deceived by various fake information. In this paper, we analyse a scenario based on an actual incident of fraud called business email compromise (BEC). We suppose that each email, in the incident, step-wisely changes user's thinking and makes him or her believe the emails. If fraud has such a step-by-step mechanism, it allows us to consider counter measures for dissuading a user from decision-making on the way. We discuss features and factors of the incident based on formulations of Channel theory. Our analysis revealed that each email message influenced user's decision-making by what kind of logical trap. It can be fundamental knowledge that is capable of warning the user with predicting logical traps. This paper contributes to providing a novel viewpoint to develop systems for detecting deception of BEC.

Speech-driven Face Reenactment for a Video Sequence Reviewed

Y. Nakashima, T. Yasui, L. Nguyen, N. Babaguchi

ITE Transactions on Media Technology and Applications   8 ( 1 )   60 - 68   2020.1

　More details

Online ISSN : 2186-7364

Anonymization of Gait Silhouette Video by Perturbing Its Phase and Shape Components Reviewed

Y. Hirose, K. Nakamura, N. Nitta, N. Babaguchi

Proc.of Asia-Pacific Signal and Processing Association Annual Summit and Conference (APSIPA ASC 2019)   2019.11