English/Japanese

椋田悠介

Curriculum Vitae (in English, last update:3/14/2024)

所属

東京大学先端科学技術研究センター
原田研究室

略歴

2013年3月
工学学士(東京大学)
2015年3月
情報理工学修士(東京大学)
2018年3月
情報理工学博士(東京大学)
2013年12月 - 2014年3月
国立天文台にてインターンシップを行いました。
2016年8月
NTTコミュニケーション科学基礎研究所にてインターンシップを行いました。
2017年4月 - 2018年3月
日本学術振興会特別研究員(DC2)
2018年4月 - 2020年4月
助教(東京大学)
2020年5月 - 現在
講師(東京大学)
2018年10月 - 現在
客員研究員(理研AIP)

連絡先

e-mail: mukuta[at]mi.t.u-tokyo.ac.jp

研究分野

画像認識,機械学習
画像特徴抽出の理論やアルゴリズムに興味があります。

論文

プレプリント

  1. Ryo Watanabe, Yusuke Mukuta, Tatsuya Harada. Fully Spiking Denoising Diffusion Implicit Models. arXiv:2312.01742, 2023.[arXiv]
  2. Ryo Umagami, Yu Ono, Yusuke Mukuta, Tatsuya Harada. HiPerformer: Hierarchically Permutation-Equivariant Transformer for Time Series Forecasting. arXiv:2305.08073, 2023.[arXiv]
  3. Yusuke Mukuta, Tatsuya Harada. Self-Supervised Learning for Group Equivariant Neural Networks. arXiv:2303.04427, 2023.[arXiv]
  4. Bumjun Jung, Yusuke Mukuta, Tatsuya Harada. Grouped self-attention mechanism for a memory-efficient Transformer. arXiv:2210.00440, 2022.[arXiv]
  5. Yusuke Mori, Hiroaki Yamane, Yusuke Mukuta, Tatsuya Harada. Computational Storytelling and Emotions: A Survey. arXiv:2205.10967, 2022.[arXiv]
  6. Ryoma Kobayashi, Yusuke Mukuta, Tatsuya Harada. Risk Consistent Multi-Class Learning from Label Proportions. arXiv:2203.12836, 2022.[arXiv]
  7. Sho Maeoki, Yusuke Mukuta, Tatsuya Harada. Video Moment Retrieval with Text Query Considering Many-to-Many Correspondence Using Potentially Relevant Pair. arXiv:2106.13566, 2021.[arXiv]
  8. Naoya Fushishita, Antonio Tejero-de-Pablos, Yusuke Mukuta, Tatsuya Harada. Efficient training for future video generation based on hierarchical disentangled representation of latent variables. arXiv:2106.03502, 2021.[arXiv]
  9. Wataru Kawai, Yusuke Mukuta, Tatsuya Harada. GRAM: Scalable Generative Models for Graphs with Graph Attention Mechanism. arXiv:1906.01861, 2019.[arXiv]
  10. Yusuke Mukuta, Tatsuaki Machida, Tatsuya Harada. Compact Approximation for Polynomial of Covariance Feature. arXiv:1906.01851, 2019.[arXiv]
  11. Keisuke Hagiwara, Yusuke Mukuta, Tatsuya Harada. End-to-End Learning Using Cycle Consistency for Image-to-Caption Transformations. arXiv:1903.10118, 2019.[arXiv]

論文誌(査読付き)

  1. Yusuke Mukuta, Tatsuya Harada. Invariant Feature Coding using Tensor Product Representation. Transactions on Machine Learning Research, 2023.[pdf][arXiv]
  2. Yusuke Mori, Hiroaki Yamane, Ryohei Shimizu, Yusuke Mukuta, Tatsuya Harada. COMPASS: a Creative Support System that Alerts Novelists to the Unnoticed Missing Contents. Computer Speech & Language, 2022.[pdf][arXiv]
  3. Takayuki Hara, Yusuke Mukuta, Tatsuya Harada. Spherical Image Generation From a Few Normal-Field-of-View Images by Considering Scene Symmetry. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022.[pdf][arXiv]
  4. Tadafumi Takata, Yusuke Mukuta, Yoshihiko Mizumoto. Modeling the Variability of Active Galactic Nuclei by Infinite Mixture of Ornstein-Uhlenbeck(OU) Processes. Astrophysical Journal, Volume 869, Number 2, pp.178-196 2018.[pdf][arXiv]

国際学会(査読付き)

  1. Kohtaro Tanaka, Kohei Uehara, Lin Gu, Yusuke Mukuta, Tatsuya Harada. Content-Specific Humorous Image Captioning Using Incongruity Resolution Chain-of-Thought. In Findings of the Association for Computational Linguistics (NAACL 2024 Findings), accepted, 2024.
  2. Motoki Omura, Takayuki Osa, Yusuke Mukuta, Tatsuya Harada. Symmetric Q-Learning: Reducing Skewness of Bellman Error in Online Reinforcement Learning. In the 38th AAAI Conference on Artificial Intelligence (AAAI 2024), accepted, 2024.[arXiv]
  3. Kohei Shiba, Yusuke Mukuta, Tatsuya Harada. Zero-shot Object Classification with Large-scale Knowledge Graph. In the 2nd Workshop on Learning with Limited Labelled Data for Image and Video Understanding (CVPR 2023 Workshop), pp.4990-4997, 2023.[pdf]
  4. Shunsuke Takahama, Yusuke Kurose, Yusuke Mukuta, Hiroyuki Abe, Akihiko Yoshizawa, Tetsuo Ushiku, Masashi Fukayama, Masanobu Kitagawa, Masaru Kitsuregawa, Tatsuya Harada. Domain Adaptive Multiple Instance Learning for Instance-level Prediction of Pathological Images. In the 20th IEEE International Symposium on Biomedical Imaging (ISBI 2023), pp.1-5, 2023.[pdf][arXiv]
  5. Thomas Westfechtel, Hao-Wei Yeh, Qier Meng, Yusuke Mukuta, Tatsuya Harada. Backprop Induced Feature Weighting for Adversarial Domain Adaptation with Iterative Label Distribution Alignment. In the Winter Conference on Applications of Computer Vision (WACV 2023), pp.392-401, 2023.[pdf]
  6. Daiki Naruse, Tomoyuki Takahata, Yusuke Mukuta, Tatsuya Harada. Pop Music Generation with Controllable Phase Length. In the 23rd International Society for Music Information Retrieval Conference (ISMIR 2022), pp.125-131, 2022.[pdf]
  7. Kohtaro Tanaka, Hiroaki Yamane, Yusuke Mori, Yusuke Mukuta, Tatsuya Harada. Learning to Evaluate Humor in Memes Based on the Incongruity Theory. In When creative AI meets conversational AI (COLING 2022, Workshop), pp.81-93, 2022.[pdf]
  8. Yuki Kawana, Yusuke Mukuta, Tatsuya Harada. Unsupervised Pose-aware Part Decomposition for Man-made Articulated Objects. In the 17th European Conference on Computer Vision (ECCV 2022), pp.558-575, 2022.[pdf][arXiv]
  9. Kohei Uehara, Yusuke Mori, Yusuke Mukuta, Tatsuya Harada. ViNTER: Image Narrative Generation with Emotion-Arc-Aware Transformer. In the 1st International Workshop on Multimodal Understanding for the Web and Social Media (WWW 2022, Workshop), pp.716-725, 2022.[pdf][arXiv]
  10. Hiromichi Kamata, Yusuke Mukuta, Tatsuya Harada. Fully Spiking Variational Autoencoder. In the 36th AAAI Conference on Artificial Intelligence, (AAAI 2022), pp.7059-7067, 2022.[pdf][arXiv]
  11. Ayaka Ideno, Yusuke Mukuta, Tatsuya Harada. Generation of Variable-Length Time Series from Text using Dynamic Time Warping-Based Method. In the 3rd ACM International Conference on Multimedia in Asia Workshop on Multi-Modal Embedding and Understanding (ACMMM Asia 2021, Workshop), 2021.[pdf]
  12. Tomu Hirata, Yusuke Mukuta, Tatsuya Harada. Making Video Recognition Models Robust to Common Corruptions With Supervised Contrastive Learning. In the 3rd ACM International Conference on Multimedia in Asia Workshop on Visual Tasks and Challenges under Low-qualiy Multimedia Data (ACMMM Asia 2021, Workshop), 2021.[pdf]
  13. Wataru Kawai, Yusuke Mukuta, Tatsuya Harada. Real-Time Mesh Extraction from Implicit Functions via Direct Reconstruction of Decision Boundary. In 2021 IEEE International Conference on Robotics and Automation (ICRA 2021), pp.12875-12881, 2021.[pdf]
  14. Yusuke Mori, Hiroaki Yamane, Yusuke Mukuta, Tatsuya Harada. The Nectar of Missing Position Prediction for Story Completion. In the 4th International Workshop on Narrative Extraction from Texts (ECIR 2021, Workshop), accepted, 2021.
  15. Ryohei Shimizu, Yusuke Mukuta, Tatsuya Harada. Hyperbolic Neural Networks++. In the 9th International Conference on Learning Representations (ICLR 2021), 2021.[pdf][arXiv]
  16. Takayuki Hara, Yusuke Mukuta, Tatsuya Harada. Spherical Image Generation from a Single Image by Considering Scene Symmetry. In the 35th AAAI Conference on Artificial Intelligence (AAAI 2021), pp.1513-1521, 2021.[pdf][arXiv]
  17. Yusuke Mori, Hiroaki Yamane, Yusuke Mukuta, Tatsuya Harada. The Nectar of Missing Position Prediction for Story Completion. In the 4th International Workshop on Narrative Extraction from Texts (ECIR 2021, Workshop), pp.65-69, 2021.[pdf]
  18. Yuki Kawana, Yusuke Mukuta, Tatsuya Harada. Neural Star Domain as Primitive Representation. In the 34th Conference on Neural Information Processing Systems (Neurips 2020), 2020.[pdf][arXiv]
  19. Naoya Fushishita, Antonio Tejero-de-Pablos, Yusuke Mukuta, Tatsuya Harada. Long-Term Video Generation of Multiple Futures Using Human Poses. In the Advances in Image Manipulation Workshop and Challenges (ECCV 2020, Workshop), pp.596-612, 2020.[pdf][arXiv]
  20. Yusuke Mori, Hiroaki Yamane, Yusuke Mukuta, Tatsuya Harada. Toward a Better Story End: Collecting Human Evaluation with Reasons. In the 12th International Conference on Natural Language Generation (INLG 2019), pp.383-390, 2019.[pdf]
  21. Kosuke Arase, Yusuke Mukuta, Tatsuya Harada. Rethinking Task and Metrics of Instance Segmentation on 3D Point Clouds. In the 4th Geometry Meets Deep Learning Workshop (ICCV 2019, Workshop), pp.4105-4113, 2019.[pdf][arXiv]
  22. Shusuke Takahama, Yusuke Kurose, Yusuke Mukuta, Hiroyuki Abe, Masashi Fukayama, Akihiko Yoshizawa, Masanobu Kitagawa, Tatsuya Harada. Multi-Stage Pathological Image Classification Using Semantic Segmentation. In the 17th International Conference on Computer Vision (ICCV 2019), pp.10702-10711, 2019.[pdf][arXiv]
  23. Antonio Tejero-de-Pablos, Kaikai Huang, Hiroaki Yamane, Yusuke Kurose, Yusuke Mukuta, Yasushi Koyama, Junichi Iho, Youji Tokunaga, Makoto Horie, Keisuke Nishizawa, Yusaku Hayashi, Tatsuya Harada. Texture-based classification of significant stenosis in CCTA multi-view images of coronary arteries. In the 22nd International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2019), pp.732-740, 2019.[pdf]
  24. Akane Iseki, Yusuke Mukuta, Yoshitaka Ushiku, Tatsuya Harada. Estimating the Causal Effect from Partially Observed Time Series. In the 33rd AAAI Conference on Artificial Intelligence (AAAI 2019), pp.3919-3926, 2019.[pdf]
  25. Yusuke Mukuta, Yoshitaka Ushiku, Tatsuya Harada. Alternating Circulant Random Features for Semigroup Kernels. In the 32nd AAAI Conference on Artificial Intelligence (AAAI 2018), pp.3836-3843, 2018.[pdf]
  26. Yusuke Mukuta, Akisato Kimura, David Adrian, Zoubin Gharamani. Weakly Supervised Collective Feature Learning from Curated Media. In the 32nd AAAI Conference on Artificial Intelligence (AAAI 2018), pp.7260-7267, 2018.[pdf]
  27. Yusuke Mukuta, Yoshitaka Ushiku, Tatsuya Harada. Spatial-Temporal Weighted Pyramid using Spatial Orthogonal Pooling. In the 16th International Conference on Computer Vision Workshop on Compact and Efficient Feature Representation and Learning in Computer Vision, pp.1041-1049, 2017.[pdf]
  28. Kuniaki Saito, Yusuke Mukuta, Yoshitaka Ushiku and Tatsuya Harada. Deep Modality Invariant Adversarial Network for Shared Representation Learning. In the 16th International Conference on Computer Vision Workshop on Transferring and Adapting Source Knowledge in Computer Vision, pp.2623-2629, 2017.[pdf]
  29. Yusuke Mukuta and Tatsuya Harada. Kernel Approximation via Empirical Orthogonal Decomposition for Unsupervised Feature Learning. the 29th IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2016), pp.5222-5230, 2016.[pdf][supplemental pdf]
  30. Yoshitaka Ushiku, Masataka Yamaguchi, Yusuke Mukuta, and Tatsuya Harada. Common Subspace for Model and Similarity: Phrase Learning for Caption Generation from Images. the 14th International Conference on Computer Vision (ICCV 2015), pp.2668-2676, 2015.[pdf]
  31. Asako Kanezaki, Yusuke Mukuta, and Tatsuya Harada. Mirror Reflection Invariant HOG descriptors for Object Detection. the 21st IEEE International Conference on Image Processing (ICIP 2014), pp.1594-1598, 2014.[pdf]
  32. Yusuke Mukuta and Tatsuya Harada. Probabilistic Partial Canonical Correlation Analysis. the 31st International Conference on Machine Learning (ICML 2014), pp.1449-1457, 2014.[pdf][supplemental pdf][code]

国内学会(査読なし)

  1. 福原大翔, 田中康太郎, 高畑智之, 椋田悠介, 原田達也. 漫画のマルチモーダル情報を用いたセリフの話者推定. 第9回コミック工学研究会, 2023年3月.
  2. 馬上凌, 椋田悠介, 原田達也. 階層的な順序共変性を考慮した時系列予測. IBIS2022, 2022年11月.
  3. 椋田悠介, 牛久祥孝, 原田達也. 交代巡回ランダム特徴によるセミグループカーネルの高速な近似. IBISML, 2017年9月.
  4. 椋田悠介, 原田達也. 確率的偏正準相関分析. IBIS2013, 2013年11月.

招待講演

  1. Yusuke Mukuta. On training and application of equivariant neural networks. In the 7th International Workshop on Symbolic-Neural Learning (SNL2023), 2023.
  2. 椋田悠介. 幾何学的制約に基づくコンパクトな画像特徴量コーディング, PRMU, 2018年10月
  3. Yusuke Mukuta, Yoshitaka Ushiku, Tatsuya Harada. Alternating Circulant Random Features for Semigroup Kernels. the 17th Forum on information Technology (FIT 2018), 2018.
  4. Yusuke Mukuta and Tatsuya Harada. Probabilistic Partial Canonical Correlation Analysis. the 18th Meeting on Image Recognition and Understanding (MIRU 2015), 2015.

受賞歴

  1. 2015 3rd place in the object detection task with external data. Large Scale Visual Recognition Challenge 2015 (ILSVRC2015).