Ruigang GE
Organization
School of Engineering Assistant Professor
Research Fields, Keywords
大規模言語モデル, 畳み込みニューラルネットワーク, 感性工学, 転移学習, 病理画像分類
Teaching and Research Fields
When one gazes at an image, subjective sensory impressions—such as “this color palette evokes a calm, tranquil mood” or “the sharpness of the contours conveys a sense of strength”—naturally arise. In this study, we integrate the semi‑structured Evaluation Grid Interview (EGI) method with large‑scale language models to develop a framework that automatically extracts, quantifies, and visualizes the evaluative terms and comparative conditions elicited. Furthermore, by employing machine learning to associate image features—namely hue, luminance, and texture—with human subjective impression ratings, we propose a sensibility‑engineering model that encapsulates qualities such as “softness,” “cleanliness,” and “warmth,” thereby enabling targeted feedback for computer‑graphics and user‑interface design. Finally, we explore a transfer‑learning strategy in which a convolutional neural network pretrained on ImageNet for natural images is fine‑tuned for pathological image classification, achieving high‑precision tumor detection even under limited data conditions.
Mail Address
ruigangge@kwansei.ac.jp
Syllabus information URL
SDGs Related Goals
Degree 【 display / non-display 】
-
Degree name:博士（ 工学 ）
Classified degree field:Informatics / Perceptual information processing
Conferring institution:AKita Prefectural University
Acquisition way:Coursework
Date of acquisition:2025.03 -
Degree name:修士（ 工学 ）
Classified degree field:Informatics / Intelligent informatics
Conferring institution:Ashikaga Institute of Technology
Acquisition way:Coursework
Date of acquisition:2019.03
Papers 【 display / non-display 】
-
Language: English
Title: Tumor detection in breast cancer pathology patches using a Multi-scale Multi-head Self-attention Ensemble Network on Whole Slide Images
Journal name: ELSEVIER: Machine Learning with Applications vol.18
Date of publication: 2024.10
Author(s): Ge, R., Chen, G., Saruta, K., & Terata, Y.DOI: https://doi.org/10.1016/j.mlwa.2024.100592
Type of publication: Research paper (scientific journal)
-
Language: English
Title: Detection of presence or absence of metastasis in WSI patches of breast cancer using the dual-enhanced convolutional ensemble neural network
Journal name: ELSEVIER: Machine Learning with Applications vol.17
Date of publication: 2024.07
Author(s): Ge, R., Chen, G., Saruta, K., & Terata, Y.DOI: https://doi.org/10.1016/j.mlwa.2024.100579
Type of publication: Research paper (scientific journal)
-
Language: English
Title: ECA-Resunet: Lung Segmentation of CT Images Based on an Efficient Channel Attention Mechanism Deep Residual-UNet
Journal name: IIEEJ Transactions on Image Electronics and Visual Computing vol.11 (1 ) (p.13 - 21)
Date of publication: 2023.02
Author(s): Peng, J., Ge, R., Chen, G., Saruta, K., and Terata, Y.DOI: https://doi.org/10.11371/tievciieej.11.1_13
Type of publication: Research paper (scientific journal)
-
Language: English
Title: MDDCNN: Diagnosis of lymph node metastases in breast cancer based on dual-CNN fusion and segmental convolution
Journal name: Information vol.24 (2) (p.129 - 138)
Date of publication: 2021.06
Author(s): Ge, R., Chen, G., Saruta, K., & Terata, Y.Type of publication: Research paper (scientific journal)
-
Language: English
Title: MAF-Net: A Multi-Stream Attention Fusion Network Integrating VGG-16 and DenseNet-121 for Patch-Level Detection of Breast Cancer Metastasis. Proc. the 4th International Conference on Image
Journal name: Proc. the 4th International Conference on Image, Signal Processing and Pattern Recognition (ISPP 2025)
Date of publication: 2025.02
Author(s): Ge, R., Chen, G., Saruta, K., & Terata, Y.Type of publication: Research paper (other academic)
MISC 【 display / non-display 】
-
Language：Japanese
Title:Deep learning techniques for metastasis detection in breast cancer WSI pathological patches
Journal name:バイオクリニカ vol.39 (11) (p.997 - 1001)
Date of publication:2024.10
Author(s):葛睿剛, 陳国躍Type of publication：Article, review, commentary, editorial, etc. (scientific journal)
-
Language：Japanese
Title:Metastasis determination of breast cancer pathological images using AI technology
Journal name:細胞
Date of publication:2023.11
Author(s):葛睿剛, 陳国躍Type of publication：Article, review, commentary, editorial, etc. (scientific journal)
Presentations 【 display / non-display 】
-
Language:English
Conference name:Proc. the 4th International Conference on Image, Signal Processing and Pattern Recognition (ISPP 2025)
International/Domestic presentation:International presentation
Presentation date：2025
Title:MAF-Net: A Multi-Stream Attention Fusion Network Integrating VGG-16 and DenseNet-121 for Patch-Level Detection of Breast Cancer Metastasis
Presentation type:Oral presentation (general)
-
Language:English
Conference name:Proc. the 5th International Conference on Intelligent Medicine and Health (ICIMH 2024)
International/Domestic presentation:International presentation
Title:Study on Metastatic Breast Cancer Detection Using the Adaptively Weighted Ensemble Convolutional Neural Network
Presentation type:Oral presentation (general)
Preferred joint research theme 【 display / non-display 】
-
Preferred joint research theme: Visual Image Analysis Utilizing AI‑Driven Multimodal Models
Joint research form:Cooperative Research with Industry-University research organizations and private agencies.
Possible form for cooperating Industry-Academia Collaboration:Funded Research