Keisuke Maeda 研究室

主宰者：Keisuke Maeda

北海道大学

AI 要約（直近 5 年の研究成果）

Keisuke Maeda研究室は、大規模な基盤モデルと生成モデルの適応・安全性向上に関する研究を展開しています。具体的には、テキスト画像生成や音楽生成、動作生成などの生成タスクにおいて、複数の概念や新規タスクを継続的に学習する際の課題に取り組んでいます。同時に、画像分類や物体検出といった認識タスクでは、個人の好みや地理的特性といった個別性を反映したカスタマイズを実現する手法を開発しています。基盤モデルの安全性と堅牢性の強化も重要なテーマです。敵対的な攻撃から視覚言語モデルを守るために、脳活動情報やモデル間の非転移性を活用した防御手法を提案しています。さらに、分散データ環境での学習時に複数の専門的なモデルを統合するモデルマージングや、連邦学習フレームワークを用いて、プライバシーを保護しながら複数クライアントから効率的に知識を共有する仕組みを研究しています。データセット効率化も研究の柱となっており、大規模データセットを小規模で効率的な合成データに圧縮するデータセット蒸留や、生成モデルを用いた訓練データの拡張手法を開発しています。これらの研究を通じて、実運用環境での計算資源制約やプライバシー制約の下で、高性能で個別対応可能な基盤モデルの実現を目指しています。

※ AI（Claude）が、公開されている論文要旨から研究の問い・手法・主要な発見を事実情報として抽出・再構成して自動生成しています。誤りを含む可能性があるため、正確性は研究室公式情報でご確認ください。

外部リンク

研究成果（100 件）

[2026] LLM-in-the-Loop Descriptor Prompt Refinement for Dialogue-Based Big Five Personality Recognition
DOI: https://doi.org/10.1109/icct-pacific69083.2026.11518780
[2026] Foreground-Aware Dataset Distillation via Dynamic Patch Selection
[2026] Optimizing surgical strategies for osteochondritis dissecans: integrating biological and mechanical enhancements across ICRS grades
DOI: https://doi.org/10.1186/s13018-025-06594-2
[2026] Discrete Prompt Tuning via Recursive Utilization of Black-Box Multimodal Large Language Model for Personalized Visual Emotion Recognition
DOI: https://doi.org/10.1109/access.2026.3682871
[2026] Deep Generative Replay-Based Personalization With Conditional Latent Attention for Diffusion Models
DOI: https://doi.org/10.1109/access.2026.3652178
[2025] Generative Dataset Distillation Based on Self-knowledge Distillation
DOI: https://doi.org/10.1109/icassp49660.2025.10888970
[2025] Gradient-Oriented Clustered Federated Learning With Efficient Knowledge Sharing in Non-IID Settings
DOI: https://doi.org/10.1109/icassp49660.2025.10890238
[2025] Enhanced framework for generating counterfactual images with sophisticated caption and inversion-free image editing
DOI: https://doi.org/10.1117/12.3057596
[2025] Balancing generalization and personalization by sharing layers in clustered federated learning
DOI: https://doi.org/10.1117/12.3057727
[2025] [Paper] Few-shot Personalized Saliency Prediction Based on Interpersonal Gaze Patterns
DOI: https://doi.org/10.3169/mta.13.382

続きを表示（残り 90 件）

[2025] Enhancing Generative Class Incremental Learning Performance With a Model Forgetting Approach
DOI: https://doi.org/10.1109/ojsp.2025.3528900
[2025] Enhancing Classification Models With Sophisticated Counterfactual Images
DOI: https://doi.org/10.1109/ojsp.2025.3530843
[2025] Generalizing Stylized Motion Generation Method by Introducing Metadata-Independent Learning and Unified Multiple Motion Dataset
DOI: https://doi.org/10.1109/tmm.2025.3645612
[2025] Out-of-Distribution Sample Selection Generated by Diffusion Model toward Model Generalization
DOI: https://doi.org/10.1109/icip55913.2025.11084593
[2025] Context-aware Image-to-Music Generation via Bridging Modalities through Musical Captions
DOI: https://doi.org/10.1145/3746027.3755794
[2025] Foul Severity Classification from Multi-View Soccer Videos with Attention and Class Characteristics
DOI: https://doi.org/10.1109/gcce65946.2025.11274603
[2025] Rethinking Continual Learning with Pre-Trained Models: Knowledge-Preserving Approach
DOI: https://doi.org/10.1109/gcce65946.2025.11274647
[2025] Similar Image Generation via Human-like Iterative Prompt Refinement with Multimodal Large Language Model
DOI: https://doi.org/10.1109/gcce65946.2025.11275204
[2025] GeoJapan Fusion Framework: A Large Multimodal Model for Regional Remote Sensing Recognition
DOI: https://doi.org/10.3390/rs17173044
[2025] Improving robustness of CLIP by adversarial training enhanced by brain activity
DOI: https://doi.org/10.1117/12.3058026
[2025] Enhancing Adversarial Robustness of Foundation Models Without Data Centralization
DOI: https://doi.org/10.1109/icip55913.2025.11084716
[2025] Cross-domain multi-step thinking: Zero-shot fine-grained traffic sign recognition in the wild
DOI: https://doi.org/10.1016/j.knosys.2025.114172
[2025] Analysis of Model Merging for Open-Vocabulary Models with Parameter Efficient Fine-Tuning Leveraging Distributed Data
DOI: https://doi.org/10.1109/icce-taiwan66881.2025.11207863
[2025] Evaluation of Effectiveness of Adversarial Defense Model Using Only Brain Activity Information
DOI: https://doi.org/10.1109/icce-taiwan66881.2025.11207822
[2025] Personalized Visual Emotion Classification through Discrete Prompt Tuning of Large Multimodal Model
DOI: https://doi.org/10.1109/icce-taiwan66881.2025.11207879
[2025] Stabilizing Self-consuming Training Loop on Music Generation Based on Music Embeddings
DOI: https://doi.org/10.1109/icce-taiwan66881.2025.11208005
[2025] Linear Structure Analysis of Embeddings for Bias Disparity Reduction in Collaborative Filtering
DOI: https://doi.org/10.1109/tsc.2025.3570852
[2025] Corrosion Detection through Integration of Multi-resolution Semantic Segmentation
DOI: https://doi.org/10.1109/itc-cscc66376.2025.11137686
[2025] Analysis of Model Merging Methods for Continual Updating of Foundation Models in Distributed Data Settings
DOI: https://doi.org/10.3390/app15095196
[2025] Enhancing Adversarial Defense via Brain Activity Integration Without Adversarial Examples
DOI: https://doi.org/10.3390/s25092736
[2025] Triplet Synthesis for Enhancing Composed Image Retrieval via Counterfactual Image Generation
DOI: https://doi.org/10.1109/icassp49660.2025.10887762
[2025] Robust Adversarial Defense Based on Non-Transferability of Attack Across Foundation Models
DOI: https://doi.org/10.1109/icassp49660.2025.10888691
[2024] Multi-Object Editing in Personalized Text-To-Image Diffusion Model Via Segmentation Guidance
DOI: https://doi.org/10.1109/icassp48485.2024.10447048
[2024] Generative Dataset Distillation Based on Large Model Pool
DOI: https://doi.org/10.1109/gcce62371.2024.10760861
[2024] Personalized Visual Emotion Classification via In-context Learning in Multimodal LLM
DOI: https://doi.org/10.1109/gcce62371.2024.10760527
[2024] An Evaluation Metric for Single Image-to-3D Models Based on a Class Confidence Score of Object Detection Models
DOI: https://doi.org/10.1109/gcce62371.2024.10761036
[2024] Emotion-conditional Image Generation Reflecting Semantic Alignment with Text-to-Image Models
DOI: https://doi.org/10.1109/gcce62371.2024.10760298
[2024] Zero-shot Composed Video Retrieval with Projection Module Bridging Modality Gap
DOI: https://doi.org/10.1109/gcce62371.2024.10760666
[2024] Zero-shot Controllable Music Generation from Videos Using Facial Expressions
DOI: https://doi.org/10.1109/gcce62371.2024.10760865
[2024] Multimodal Adversarial Defense Trained on Features Extracted from Images and Brain Activity
DOI: https://doi.org/10.1109/gcce62371.2024.10760752
[2024] Improving Zero-shot Adversarial Robustness via Integrating Image Features of Foundation Models
DOI: https://doi.org/10.1109/gcce62371.2024.10760921
[2024] Generalizing Human Motion Style Transfer Method Based on Metadata-independent Learning
DOI: https://doi.org/10.1145/3681756.3697890
[2024] An Evaluation Metric for Single Image-to-3D Models Based on Object Detection Perspective
DOI: https://doi.org/10.1145/3681758.3697992
[2024] Damage‐level classification considering both correlation between image and text data and confidence of attention map
DOI: https://doi.org/10.1111/mice.13366
[2024] What to Do and Where to Go Next? Action Prediction in Soccer Using Multimodal Co-Attention Transformer
DOI: https://doi.org/10.1145/3689061.3689076
[2024] Cross-Domain Few-Shot In-Context Learning For Enhancing Traffic Sign Recognition
DOI: https://doi.org/10.1109/icip51287.2024.10647129
[2024] Motion-STUDiO : Motion Style Transfer Utilized for Dancing Operation by Considering Both Style and Dance Features
DOI: https://doi.org/10.1109/icce-taiwan62264.2024.10674583
[2024] Zero-shot High-risk Situation Detection Based on Semantic Segmentation and Pose Estimation Using Fixed-point Cameras at Construction Sites
DOI: https://doi.org/10.1109/itc-cscc62988.2024.10628198
[2024] Generative Dataset Distillation: Balancing Global Structure and Local Details
DOI: https://doi.org/10.1109/cvprw63382.2024.00762
[2024] Multi-Modal Gaussian Process Latent Variable Model With Semi-Supervised Label Dequantization
DOI: https://doi.org/10.1109/access.2024.3437328
[2024] Medial cortical bone thickness of the tibial diaphysis in osteoarthritis is related to lower extremity alignment and tibial morphology
DOI: https://doi.org/10.1186/s13018-024-04849-y
[2024] Multimodal Shot Prediction Based on Spatial-Temporal Interaction between Players in Soccer Videos
DOI: https://doi.org/10.3390/app14114847
[2024] Multimodal Transformer Model Using Time-Series Data to Classify Winter Road Surface Conditions
DOI: https://doi.org/10.3390/s24113440
[2024] Trial Analysis of Brain Activity Information for the Presymptomatic Disease Detection of Rheumatoid Arthritis
DOI: https://doi.org/10.3390/bioengineering11060523
[2024] Algal Bed Region Segmentation Based on a ViT Adapter Using Aerial Images for Estimating CO2 Absorption Capacity
DOI: https://doi.org/10.3390/rs16101742
[2024] Analysis of Continual Learning Techniques for Image Generative Models with Learned Class Information Management
DOI: https://doi.org/10.3390/s24103087
[2024] A Novel Frame-Selection Metric for Video Inpainting to Enhance Urban Feature Extraction
DOI: https://doi.org/10.3390/s24103035
[2024] Masked modeling-based action event prediction considering bidirectional time-series in soccer
DOI: https://doi.org/10.1117/12.3019688
[2024] Automatic Findings Generation for Distress Images Using In-Context Few-Shot Learning of Visual Language Model Based on Image Similarity and Text Diversity
DOI: https://doi.org/10.20965/jrm.2024.p0353
[2024] MLLM-based Automatic Exploration of Editing Prompt for High Engagement Image Generation
DOI: https://doi.org/10.1109/gcce62371.2024.10760859
[2024] AUTOMATIC RECOGNITION OF ALGAL BED AREAS BASED ON A LARGE-SCALE SEMANTIC SEGMENTATION MODEL FOR ESTIMATING CO<sub>2</sub> ABSORPTION BY BLUE CARBON
DOI: https://doi.org/10.2208/jscejj.24-17286
[2024] Enhancing Noisy Label Learning Via Unsupervised Contrastive Loss with Label Correction Based on Prior Knowledge
DOI: https://doi.org/10.1109/icassp48485.2024.10446840
[2024] Privacy Preserving Gaze Estimation Via Federated Learning Adapted To Egocentric Video
DOI: https://doi.org/10.1109/icassp48485.2024.10446696
[2024] Caption Unification for Multi-View Lifelogging Images Based on In-Context Learning with Heterogeneous Semantic Contents
DOI: https://doi.org/10.1109/icassp48485.2024.10445969
[2024] Text-Guided Image Editing Based on Post Score for Gaining Attention on Social Media
DOI: https://doi.org/10.3390/s24030921
[2024] Individual Persistence Adaptation for User-Centric Evaluation of User Satisfaction in Recommender Systems
DOI: https://doi.org/10.1109/access.2024.3360693
[2023] TolerantGAN: Text-guided Image Manipulation Tolerant to Real-world Image
DOI: https://doi.org/10.1109/ojsp.2023.3343335
[2023] SpectralMAP: Approximating Data Manifold With Spectral Decomposition
DOI: https://doi.org/10.1109/access.2023.3257427
[2023] Material Compound-Property Retrieval Using Electron Microscope Images for Rubber Material Development
DOI: https://doi.org/10.1109/access.2023.3304341
[2023] Zero-Shot Visual Sentiment Prediction via Cross-Domain Knowledge Distillation
DOI: https://doi.org/10.1109/ojsp.2023.3344079
[2023] Flexibly manipulating popularity bias for tackling trade-offs in recommendation
DOI: https://doi.org/10.1016/j.ipm.2023.103606
[2023] Zero-Shot Traffic Sign Recognition Based on Midlevel Feature Matching
DOI: https://doi.org/10.3390/s23239607
[2023] Manipulation Direction: Evaluating Text-Guided Image Manipulation Based on Similarity between Changes in Image and Text Modalities
DOI: https://doi.org/10.3390/s23229287
[2023] Personalized Content Recommender System via Non-verbal Interaction Using Face Mesh and Facial Expression
DOI: https://doi.org/10.1145/3581783.3612675
[2023] Visual Emotion Recognition Through Multimodal Cyclic-Label Dequantized Gaussian Process Latent Variable Model
DOI: https://doi.org/10.20965/jrm.2023.p1321
[2023] Video-Music Retrieval with Fine-Grained Cross-Modal Alignment
DOI: https://doi.org/10.1109/icip49359.2023.10222739
[2023] Multi-View Variational Recurrent Neural Network for Human Emotion Recognition Using Multi-Modal Biological Signals
DOI: https://doi.org/10.1109/icip49359.2023.10222460
[2023] Text-to-image Diffusion Model Suppressing Catastrophic Forgetting via Elastic Weight Consolidation
DOI: https://doi.org/10.1109/gcce59613.2023.10315296
[2023] Caption Unification for Multiple Viewpoint Lifelogging Images and Its Verification
DOI: https://doi.org/10.1109/gcce59613.2023.10315286
[2023] Deterioration Level Estimation for Infrastructures Considering Noisy Labels via DivideMix
DOI: https://doi.org/10.1109/gcce59613.2023.10315319
[2023] Novel Feature Extraction for Classification of Auditory-visual Stimuli from fNIRS Signals
DOI: https://doi.org/10.1109/gcce59613.2023.10315546
[2023] A Controllable Recoloring Method for Novel Views Using Segment Anything Model
DOI: https://doi.org/10.1109/gcce59613.2023.10315376
[2023] Improving Visual Counterfactual Explanation Models for Image Classification via CLIP
DOI: https://doi.org/10.1109/gcce59613.2023.10315277
[2023] Text-Guided Facial Image Manipulation for Wild Images via Manipulation Direction-Based Loss
DOI: https://doi.org/10.1109/icip49359.2023.10222231
[2023] Feature Integration via Back-Projection Ordering Multi-Modal Gaussian Process Latent Variable Model for Rating Prediction
DOI: https://doi.org/10.1109/icip49359.2023.10222279
[2023] Zero-Shot Neural Decoding with Semi-Supervised Multi-View Embedding
DOI: https://doi.org/10.3390/s23156903
[2023] Shoot Event Prediction in Soccer Considering Expected Goals Based on Players’ Positions
DOI: https://doi.org/10.1109/icce-taiwan58799.2023.10226874
[2023] Estimation of Amyloid-β Positivity Using QSM Images Considering Age Information
DOI: https://doi.org/10.1109/icce-taiwan58799.2023.10226714
[2023] Prediction of Shoot Events by Considering Spatio-temporal Relations of Multimodal Features
DOI: https://doi.org/10.1109/icce-taiwan58799.2023.10226862
[2023] A Gaussian Process Decoder with Spectral Mixtures and a Locally Estimated Manifold for Data Visualization
DOI: https://doi.org/10.3390/app13148018
[2023] Multi-Label Classification in Anime Illustrations Based on Hierarchical Attribute Relationships
DOI: https://doi.org/10.3390/s23104798
[2023] Gaze-Dependent Image Re-Ranking Technique for Enhancing Content-Based Image Retrieval
DOI: https://doi.org/10.3390/app13105948
[2023] Prediction of Shooting Events in Soccer Videos Using Complete Bipartite Graphs and Players’ Spatial-Temporal Relations
DOI: https://doi.org/10.3390/s23094506
[2023] Estimation of Visual Contents from Human Brain Signals via VQA Based on Brain-Specific Attention
DOI: https://doi.org/10.1109/icassp49357.2023.10096890
[2023] 画像の感情推定のためのラベル逆量子化を導入した正準相関分析
[2023] Estimation of Degradation Degree in Road Infrastructure Based on Multi-Modal ABN Using Contrastive Learning
DOI: https://doi.org/10.3390/s23031657
[2023] Hierarchical Multi-Label Attribute Classification With Graph Convolutional Networks on Anime Illustration
DOI: https://doi.org/10.1109/access.2023.3265728
[2023] Text-Guided Image Manipulation via Generative Adversarial Network With Referring Image Segmentation-Based Guidance
DOI: https://doi.org/10.1109/access.2023.3269847
[2022] Affective Embedding Framework with Semantic Representations from Tweets for Zero-Shot Visual Sentiment Prediction
DOI: https://doi.org/10.1145/3551626.3564950
[2022] Popularity-Aware Graph Social Recommendation for Fully Non-Interaction Users
DOI: https://doi.org/10.1145/3551626.3564969

科研費（0 件）

まだデータがありません（KAKEN 取り込み後に表示）。

所属学会・役職（0 件）

まだデータがありません（学会データ連携後に表示）。

AI 要約（直近 5 年の研究成果）

外部リンク

関連研究室(8 件)

研究成果（100 件）

科研費（0 件）

所属学会・役職（0 件）