Tomoya Sakai
Profile
Publications
Software
Misc
（日本語ページ）

Publications

Journal Articles

  1. T. Sakai, G. Niu, & M. Sugiyama
    Information-theoretic representation learning for positive-unlabeled classification.
    Neural Computation, vol.33, no.1, pp.244-268, 2021.
    [paper]

  2. H. Bao, T. Sakai, I. Sato, & M. Sugiyama
    Convex formulation of multiple instance learning from positive and unlabeled bags.
    Neural Networks, vol.105, pp.132-141, 2018.
    [paper] [preprint]

  3. T. Sakai, G. Niu, & M. Sugiyama
    Semi-supervised AUC optimization based on positive-unlabeled learning.
    Machine Learning, vol.107, no.4, pp.767-794, 2018.
    (Presented at Asian Conference on Machine Learning (ACML 2017), Seoul, Korea, 15-17, Nov., 2017)
    [paper] [preprint] [code (MATLAB)]

  4. T. Sakai, M. Sugiyama, K. Kitagawa, & K. Suzuki
    Registration of infrared transmission images using squared-loss mutual information.
    Precision Engineering, vol.39, pp.187-193, 2015.
    [paper]

  5. T. Sakai & M. Sugiyama
    Computationally efficient estimation of squared-loss mutual information with multiplicative kernel models.
    IEICE Transactions on Information and Systems, vol.E97-D, no.4, pp.968-971, 2014.
    [paper] [code (MATLAB)]

Conference Papers (full review)

  1. T. Sakai, H. Qiu, T. Katsuki, D. Kimura, T. Osogami, & T. Inoue.
    A surprisingly simple approach to generalized few-shot semantic segmentation.
    In Advances in Neural Information Processing Systems 37 (NeurIPS 2024), pp. xxxx-xxxx, 2024.
    [To appear]

  2. T. Sakai
    A generalized backward compatibility metric.
    In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD ’22), pp. 1525–1535, 2022.
    (Presented at the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD ’22), Washington, DC, USA, August 14–18, 2022)
    (Presented by Ryuta Matsuno)
    [paper]

  3. T. Sakai
    Source hypothesis transfer for zero-shot domain adaptation.
    In Machine Learning and Knowledge Discovery in Databases. Research Track, pp. 570–586, 2021.
    (Presented at the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD 2021), Virtual, Sep. 13-17, 2021)
    [paper] [pdf]

  4. A. Tanimoto, T. Sakai, T. Takenouchi, & H. Kashima
    Causal combinatorial factorization machines for set-wise recommendation.
    In Advances in Knowledge Discovery and Data Mining, pp. 498–509, 2021.
    (Presented at the 25th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD 2021), Virtual, May 11-14, 2021)
    [paper]

  5. A. Tanimoto, T. Sakai, T. Takenouchi, & H. Kashima
    Regret minimization for causal inference on large treatment space.
    In Proceedings of the 24th International Conference on Artificial Intelligence and Statistics (AISTATS 2021), 2021.
    (Presented at the 24th International Conference on Artificial Intelligence and Statistics, Virtual, April 13-15, 2021)
    [paper]

  6. T. Sakai & N. Ohsaka
    Predictive optimization with zero-shot domain adaptation.
    In Proceedings of the 2021 SIAM International Conference on Data Mining (SDM21), 2021.
    (Presented at the 2021 SIAM International Conference on Data Mining (SDM21), Virtual, April 29 - May 1, 2021)
    [paper] [preprint]

  7. T. Ishida, I. Yamane, T. Sakai, G. Niu, & M. Sugiyama
    Do we need zero training loss after achieving zero training error?
    In Proceedings of the Thirty-seventh International Conference on Machine Learning, PMLR, vol. 119, pp. 4604-4614, 2020.
    (Presented at Thirty-seventh International Conference on Machine Learning (ICML2020), Virtual, July 13-18, 2020)
    [paper] [preprint] [code (Python)]

  8. H. Sasaki, T. Sakai, & T. Kanamori
    Robust modal regression with direct log-density derivative estimation.
    In Proceedings of the 35th Conference on Uncertainty in Artificial Intelligence, PMLR, vol. 124, pp. 380-389, 2020.
    (Presented at Conference on Uncertainty in Artificial Intelligence (UAI2020), Virtual, Aug. 3-6, 2020)
    [paper] [preprint]

  9. N. Ohsaka, T. Sakai, & A. Yabe
    A predictive optimization framework for hierarchical demand matching.
    In Proceedings of the 2020 SIAM International Conference on Data Mining (SDM20), pp. 172-180, 2020.
    [paper]

  10. T. Sakai & N. Shimizu
    Covariate shift adaptation on learning from positive and unlabeled data.
    In Proceedings of the Thirty-third AAAI Conference on Artificial Intelligence, vol. 33, no. 1, pp. 4838-4845, 2019.
    (Presented at AAAI Conference on Artificial Intelligence (AAAI-19), Hawaii, USA, Jan. 27 - Feb 1, 2019)
    [paper]

  11. T. Sakai, M. C. du Plessis, G. Niu, & M. Sugiyama
    Semi-supervised classification based on classification from positive and unlabeled data.
    In D. Precup, Y.-W. Teh (Eds.), Proceedings of The 34th International Conference on Machine Learning (ICML 2017), vol. 70, pp. 2998-3006, 2017.
    (Presented at International Conference on Machine Learning (ICML 2017), Sydney, Australia, Aug. 6-11, 2017)
    [paper] [code (MATLAB)] [code (Python)]

  12. M. Ashizawa, H. Sasaki, T. Sakai, & M. Sugiyama
    Least-squares log-density gradient clustering for Riemannian manifolds.
    In A. Singh and J. Zhu (Eds.), Proceedings of the 20th International Conference on Artificial Intelligence and Statistics, vol. 54, pp. 537-546, 2017.
    (Presented at International Conference on Artificial Intelligence and Statistics (AISTATS 2017), Fort Lauderdale, USA, April 20-22, 2017)
    [paper] [code (MATLAB)]

  13. G. Niu, M. C. du Plessis, T. Sakai, Y. Ma, & M. Sugiyama
    Theoretical comparisons of positive-unlabeled learning against positive-negative learning.
    In D. D. Lee, M. Sugiyama, U. von Luxburg, I. Guyon, & R. Garnett (Eds.), Advances in Neural Information Processing Systems 29, pp. 1199-1207, 2016.
    (Presented at Neural Information Processing Systems (NeurIPS2016), Barcelona, Spain, Dec. 5-8, 2016)
    [paper]

Workshops/Technical Reports/Demos/Domestic Conferences

  1. T. Katsuki, H. Qiu, T. Sakai, T. Osogami, T. Inoue, D. Kimura, & C. Malossi.
    Unsupervised Domain Adaptation from a Few Target Samples.
    Presented at The 27th Information-Based Induction Sciences Workshop (IBIS2024), Saitama + Online, Japan, Nov.4-Nov.7, 2024.

  2. T. Katsuki, H. Qiu, T. Sakai, T. Osogami, T. Inoue, D. Kimura, & C. Malossi.
    Domain Adaptation from Sequentially Arriving Target Images with Online Deep Sets.
    Presented at The 27th Meeting on Image Recognition and Understanding (MIRU2024), Kumamoto, Japan, Aug. 6-9, 2024.
    Selected as oral presentation

  3. T. Katsuki, H. Qiu, T. Sakai, T. Osogami, T. Inoue, D. Kimura, C. Malossi.
    Domain-adaptive Instance Segmentation for Visual Inspection from Civil Infrastructure Images.
    Presented at The 27th Meeting on Image Recognition and Understanding (MIRU2024), Kumamoto, Japan, Aug. 6-9, 2024.

  4. T. Sakai, H. Qiu, T. Katsuki, D. Kimura, T. Osogami, & T. Inoue.
    Simple Class Relation Helps Generalized Few-Shot Semantic Segmentation.
    Presented at The 27th Meeting on Image Recognition and Understanding (MIRU2024), Kumamoto, Japan, Aug. 6-9, 2024.
    Selected as oral presentation

  5. T. Frick, C. Skura, F. M. Janicki, R. Assaf, N. Avogaro, D. Caraballo, Y. G. Cinar, B. Ebouky, I. Giurgiu, T. Katsuki, P. Kluska, C. Malossi, H. Qiu, T. Sakai, F. Scheidegger, A. Simeski, D. Yang, A. Bartezzaghi, M. Rigotti.
    Probabilistic Feature Matching for Fast Scalable Visual Prompting.
    Demo at The 33rd International Joint Conference on Artificial Intelligence (IJCAI2024), Jeju, Korea, Aug. 3-9, 2024.
    [link1] [link2]

  6. T. Frick, C. Skura, F. M. Janicki, R. Assaf, N. Avogaro, D. Caraballo, Y. G. Cinar, B. Ebouky, I. Giurgiu, T. Katsuki, P. Kluska, C. Malossi, H. Qiu, T. Sakai, F. Scheidegger, A. Simeski, D. Yang, A. Bartezzaghi, & M. Rigotti.
    Interactive Image Segmentation Guided by Visual Prompting.
    Demo at The IEEE/CVF Conference on Computer Vision and Pattern Recognition 2024 (CVPR2024), Washington, USA, June 17-21, 2024.
    [link1] [link2]

  7. H. Qui, T. Katsuki, T. Sakai, T. Osogami, D. Kimura & T. Inoue
    Crack Detection by Cycle-Consistency.
    Presented at The 30th Symposium on Sensing via Image Information (SSII2024), Kanagawa, Japan, June 12-14, 2024.
    [link (in Japanese)]

  8. T. Sakai, T. Katsuki, H. Qui, T. Osogami, & T. Inoue
    Preventing Catastrophic Forgetting in Generalized Few-Shot Semantic Segmentation.
    Proceedings of the 38th Annual Conference of the Japanese Society for Artificial Intelligence, 2024.
    (Presented at the 38th Annual Conference of the Japanese Society for Artificial Intelligence (JSAI2024), Shizuoka + Online, Japan, May 28-31, 2024)
    [paper (in Japanese)]
    Best Interactive Session Award

  9. T. Sakai, T. Katsuki, H. Qui, T. Osogami, & T. Inoue
    Simple Yet Effective Approach to Generalized Few-Shot Semantic Segmentation.
    Presented at The 26th Information-Based Induction Sciences Workshop (IBIS2023), Fukuoka + Online, Japan, Oct.29-Nov.1, 2023.

  10. K. Sakuma, T. Sakai, & Y. Kameda
    A Method of Identifying Causes of Prediction Errors to Accelerate MLOps.
    Proceedings of the 35th Annual Conference of the Japanese Society for Artificial Intelligence, 2021.
    (Presented at the 35th Annual Conference of the Japanese Society for Artificial Intelligence (JSAI2021), Virtual, June 8-11, 2021)
    [paper (in Japanese)]

  11. T. Ishida, I. Yamane, T. Sakai, G. Niu, & M. Sugiyama
    Do we need zero training loss after achieving zero training error?
    Presented at Virtual at The 23rd Information-Based Induction Sciences Workshop (IBIS2020), Virtual, Nov. 23-26, 2020.
    Finalist of Outstanding Presentation Award

  12. T. Sakai & N. Ohsaka
    Predictive optimization based on zero-shot domain adaptation.
    Proceedings of the 34th Annual Conference of the Japanese Society for Artificial Intelligence, 2020.
    (Presented at the 34th Annual Conference of the Japanese Society for Artificial Intelligence (JSAI2020), Virtual, June 9-12, 2020)
    [paper (in Japanese)]

  13. Y. Sogawa & T. Sakai
    Semi-supervised domain adaptation using prediction models in associated domains.
    Proceedings of the 33rd Annual Conference of the Japanese Society for Artificial Intelligence, 2019.
    (Presented at the 33th Annual Conference of the Japanese Society for Artificial Intelligence (JSAI2019), Niigata, Japan, June 4-7, 2019)
    [paper (in Japanese)]

  14. T. Sakai & Y. Sogawa
    An approach to unseen classes classification with in-service predictors.
    Proceedings of the 33rd Annual Conference of the Japanese Society for Artificial Intelligence, 2019.
    (Presented at the 33th Annual Conference of the Japanese Society for Artificial Intelligence (JSAI2019), Niigata, Japan, June 4-7, 2019)
    [paper (in Japanese)]
    Best Interactive Session Award

  15. T. Sakai, G. Niu, & M. Sugiyama
    Semi-supervised AUC optimization based on positive-unlabeled learning.
    Presented at ERATO Winter Festa Episode 3, Tokyo, Japan, Dec. 25-26, 2017.

  16. T. Sakai, G. Niu, & M. Sugiyama
    Semi-supervised AUC optimization based on positive-unlabeled learning.
    IEICE Technical Report, IBISML2017-40, vol. 117, no. 293, pp. 39-46, Nov., 2017.
    (Presented at The 20th Information-Based Induction Science Workshop (IBIS2017), Tokyo, Japan, Nov. 8-11, 2017)

  17. T. Sakai, M. C. du Plessis, G. Niu, & M. Sugiyama
    Semi-supervised classification based on positive-unlabeled classification.
    (Presented at First International Workshop on Symbolic-Neural Learning (SNL-2017), Nagoya, Japan, July 7-8, 2017.)

  18. H. Bao, T. Sakai, M. Sugiyama, & I. Sato
    Risk minimization framework for multiple instance learning from positive and unlabeled bags.
    (Presented at First International Workshop on Symbolic-Neural Learning (SNL-2017), Nagoya, Japan, July 7-8, 2017.)

  19. T. Sakai, M. C. du Plessis, G. Niu, & M. Sugiyama
    Semi-supervised classification based on positive-unlabeled classification.
    Presented at The Machine Learning Summer School (MLSS 2017 Tuebingen), Tuebingen, Germany, June 19-30, 2017.

  20. G. Niu (Presented by T. Sakai)
    Positive-unlabeled learning with application to semi-supervised learning.
    Presented at Microsoft Research Asia Academic Day 2017, Yilan, Taiwan, May 26, 2017.

  21. T. Sakai, M. C. du Plessis, G. Niu, & M. Sugiyama
    Semi-supervised classification based on classification from positive and unlabeled data.
    IEICE Technical Report, IBISML2016-80, pp.243-250, Kyoto, Japan, Nov. 16-19, 2016.
    (Presented at 2016 Workshop on Information-Based Induction Sciences (IBIS2016), Kyoto, Japan, Nov. 16-19, 2016)

  22. M. Ashizawa, H. Sasaki, T. Sakai, & M. Sugiyama
    Least-squares log-density gradient clustering for Riemannian manifolds.
    IEICE Technical Report, IBISML2015-96, pp.17-24, Tokyo, Japan, Mar. 17-18, 2016.

  23. T. Sakai & M. Sugiyama
    Estimation of squared-loss mutual information and its application in registration of infrared-transmission images.
    Presented at Machine Learning Summer School 2015 Kyoto (MLSS2015 Kyoto), no.27-11, Aug. 23-Sep. 4, 2015.

  24. T. Sakai & M. Sugiyama
    Image registration using squared-loss mutual information.
    Presented at 2014 Workshop on Information-Based Induction Sciences (IBIS2014), Nagoya, Japan, Nov. 16-19, 2014.

  25. T. Sakai, M. Sugiyama, K. Kitagawa, & K. Suzuki
    Registration of infrared-transmission images using squared-loss mutual information.
    In Proceedings of the Japan Society for Precision Engineering 2014 Spring Meeting, pp.973-974, Tokyo, Japan, Mar. 18-20, 2014.
    (Presented at the Japan Society for Precision Engineering 2014 Spring Meeting, Tokyo, Japan, Mar. 18-20, 2014)

  26. T. Sakai & M. Sugiyama
    Computationally efficient estimation of squared-loss mutual information with multiplicative kernel models.
    IEICE Technical Report, IBISML2013-53, pp.131-137, Tokyo, Japan, Nov. 10-13, 2013.
    (Presented at 2013 Workshop on Information-Based Induction Sciences (IBIS2013),Tokyo, Japan, Nov. 10-13, 2013)

Books

  1. M. Sugiyama, H. Bao, T. Ishida, N. Lu, T. Sakai, & G. Niu.
    Machine learning from weak supervision: An empirical risk minimization approach,
    MIT Press, Cambridge, MA, USA, 2022.
    [link] [Amazon] [Preview by Google Books]