Projects funded by the NCN

Information on the principal investigator and host institution

Information of the project and the call

Keywords

Equipment

Delete all

Deep Generative View on Continual Learning

2020/39/B/ST6/01511

Keywords:

machine learning deep neural networks generative models continual learning computer vision

Descriptors:

  • ST6_011:
  • ST6_007:

Panel:

ST6 - Computer science and informatics: informatics and information systems, computer science, scientific computing, intelligent systems

Host institution :

Politechnika Warszawska, Wydział Elektroniki i Technik Informacyjnych

woj. mazowieckie

Other projects carried out by the institution 

Principal investigator (from the host institution):

prof. Tomasz Trzciński 

Number of co-investigators in the project: 5

Call: OPUS 20 - announced on 2020-09-15

Amount awarded: 1 316 000 PLN

Project start date (Y-m-d): 2021-07-23

Project end date (Y-m-d): 2024-11-22

Project duration:: 40 months (the same as in the proposal)

Project status: Project settled

Project description

Download the project description in a pdf file

Note - project descriptions were prepared by the authors of the applications themselves and placed in the system in an unchanged form.

Information in the final report

  • Publication in academic press/journals (3)
  • Articles in post-conference publications (16)
  1. Enhancing variational quantum state diagonalization using reinforcement learning techniques
    Authors:
    Akash Kundu, Przemysław Bedełek, Mateusz Ostaszewski, Onur Danaci, Yash J Patel, Vedran Dunjko, Jarosław A Miszczak
    Academic press:
    New Journal of Physics (rok: 2024, tom: 26, strony: 13034), Wydawca: IOP Publishing
    Status:
    Published
    DOI:
    10.1088/1367-2630/ad1b7f - link to the publication
  2. Logarithmic Continual Learning
    Authors:
    W. Masarczyk, P. Wawrzyński, D. Marczak, K. Deja, T. Trzciński
    Academic press:
    IEEE Access (rok: 2022, tom: 10, strony: 117001-117010), Wydawca: IEEE
    Status:
    Published
    DOI:
    10.1109/ACCESS.2022.3218907 - link to the publication
  3. Looking through the past: Better knowledge retention for generative replay in continual learning
    Authors:
    V. Khan, S. Cygert, K. Deja, T. Trzcinski, and B. Twardowski
    Academic press:
    IEEE Access (rok: 2024, tom: 12, strony: 45309–45317), Wydawca: Institute of Electrical and Electronics Engineers (IEEE)
    Status:
    Published
    DOI:
    10.1109/ACCESS.2024.3379148 - link to the publication
  1. The tunnel effect: Building data representations in deep neural networks
    Authors:
    Wojciech Masarczyk, Mateusz Ostaszewski, Ehsan Imani, Razvan Pascanu, Piotr Miłoś, Tomasz Trzcinski
    Conference:
    Advances in Neural Information Processing Systems 36 (rok: 2024, tom: Neural Information Processing Systems (NeurIPS), strony: b.d.), Wydawca: Curran Associates
    Data:
    konferencja 8.12.2023 - 12.12.2023
    Status:
    Published
  2. Magmax: Leveraging model merging for seamless continual learning
    Authors:
    D. Marczak, B. Twardowski, T. Trzci'nski, and S. Cygert,
    Conference:
    18th European Conference, Milan, Italy, September 29–October 4, 2024, Proceedings, Part LXII (rok: 2024, tom: ECCV: European Conference on Computer Vision, strony: 379–395), Wydawca: Springer
    Data:
    konferencja 2024-09-29 -- 2024-10-04
    Status:
    Published
  3. Adapt your teacher: Improving knowledge distillation for exemplar-free continual learning
    Authors:
    Filip Szatkowski, Mateusz Pyla, Marcin Przewięźlikowski, Sebastian Cygert, Bartłomiej Twardowski, Tomasz Trzciński
    Conference:
    Proceeedins of the IEEE/CVF International Conference on Computer Vision Workshops (ICCVW 2023) (rok: 2023, tom: International Conference on Computer Vision, strony: 3504-3509), Wydawca: IEEE
    Data:
    konferencja 2023-10-02 -- 2023-10-06
    Status:
    Published
    DOI:
    10.1109/ICCVW60793.2023.00377 - link to the publication
  4. Overestimation, Overfitting, and Plasticity in Actor-Critic: the Bitter Lesson of Reinforcement Learning
    Authors:
    Michal Nauman, Michał Bortkiewicz, Piotr Miłoś, Tomasz Trzcinski, Mateusz Ostaszewski, Marek Cygan
    Conference:
    Proceedings of Machine Learning Research (rok: 2024, tom: International Conference on Machine Learning, strony: 37342-37364), Wydawca: Brak
    Data:
    konferencja 2024-07-21 -- 2024-07-27
    Status:
    Published
  5. The Effectiveness of World Models for Continual Reinforcement Learning
    Authors:
    Samuel Kessler, Mateusz Ostaszewski, MichałPaweł Bortkiewicz, Mateusz Żarski, Maciej Wolczyk, Jack Parker-Holder, Stephen J. Roberts, Piotr Miłoś
    Conference:
    Proceedings of The 2nd Conference on Lifelong Learning Agents (rok: 2023, tom: Conference on Lifelong Learning Agents, strony: 184--204), Wydawca: PMLR
    Data:
    konferencja 22--25 Aug 2023
    Status:
    Published
  6. Revisiting supervision for continual repre- sentation learning
    Authors:
    D. Marczak, S. Cygert, T. Trzci'nski, and B. Twardowski,
    Conference:
    18th European Conference, Milan, Italy, September 29–October 4, 2024, Proceedings, Part LXII (rok: 2024, tom: ECCV: European Conference on Computer Vision, strony: 181-197), Wydawca: Springer
    Data:
    konferencja 2024-09-29 -- 2024-10-04
    Status:
    Published
    DOI:
    10.1007/978-3-031-72658-3_11 - link to the publication
  7. Category adaptation meets projected distillation in generalized continual category discovery
    Authors:
    G. Rypeść, D. Marczak, S. Cygert, T. Trzciński, and B. Twardowski,
    Conference:
    18th European Conference, Milan, Italy, September 29–October 4, 2024, Proceedings, Part LXII (rok: 2024, tom: ECCV: European Conference on Computer Vision, strony: 320-327), Wydawca: Springer
    Data:
    konferencja 2024-09-29 -- 2024-10-04
    Status:
    Published
  8. Continual Learning with Guarantees via Weight Interval Constraints
    Authors:
    M. Wołczyk, K. Piczak, B. Wójcik, Ł. Pustelnik, P. Morawiecki, J. Tabor, T. Trzcinski, P. Spurek
    Conference:
    Proceedings of the 39th International Conference on Machine Learning (rok: 2022, tom: International Conference on Machine Learning (ICML), strony: 23897-23911), Wydawca: PMLR
    Data:
    konferencja Jul 17, 2022 – Jul 23, 2022
    Status:
    Published
  9. Fine-tuning reinforcement learning models is secretly a forgetting mitigation problem
    Authors:
    M. Wolczyk, B. Cupiał, M. Ostaszewski, M. Bortkiewicz, M. Zając, R. Pascanu, Ł. Kuciński, and P. Miłoś
    Conference:
    Proceedings of Machine Learning Research (rok: 2024, tom: International Conference on Machine Learning, strony: 53039-53078), Wydawca: Brak
    Data:
    konferencja 2024-07-21 -- 2024-07-27
    Status:
    Published
  10. Curriculum reinforcement learning for quantum architecture search under hardware error
    Authors:
    Y. J. Patel, A. Kundu, M. Ostaszewski, X. Bonet-Monroig, V. Dunjko, and O. Danaci
    Conference:
    Brak (rok: 2024, tom: International Conference on Learning Representations (ICLR), strony: Brak), Wydawca: Brak
    Data:
    konferencja 2024-05-07 -- 2024-05-11
    Status:
    Published
  11. Divide and not forget: Ensemble of selectively trained experts in continual learning
    Authors:
    G. Rypeść, S. Cygert, V. Khan, T. Trzciński, B. Zieliński, and B. Twardowski
    Conference:
    Brak (rok: 2024, tom: International Conference on Learning Representations (ICLR), strony: Brak), Wydawca: Brak
    Data:
    konferencja 2024-05-07 -- 2024-05-11
    Status:
    Published
  12. Learning Data Representations with Joint Diffusion Models
    Authors:
    Kamil Deja, Tomasz Trzciński, Jakub M Tomczak
    Conference:
    Machine Learning and Knowledge Discovery in Databases: Research Track (rok: 2023, tom: European Conference, ECML PKDD 2023, strony: 543–559), Wydawca: Springer Cham
    Data:
    konferencja September 18–22, 2023
    Status:
    Published
    DOI:
    10.1007/978-3-031-43415-0 - link to the publication
  13. On robustness of generative representations against catastrophic forgetting
    Authors:
    W. Masarczyk, K. Deja, T. Trzciński
    Conference:
    Communications in Computer and Information Science, vol 1517 (rok: 2021, tom: International Conference on Neural Information Processing, strony: 325-333), Wydawca: Springer International Publishing
    Data:
    konferencja 8-12 December 2021
    Status:
    Published
    DOI:
    10.1007/978-3-030-92310-5_38 - link to the publication
  14. Multiband VAE: Latent Space Alignment for Knowledge Consolidation in Continual Learning
    Authors:
    K. Deja, P. Wawrzyński, W. Masarczyk, D. Marczak, T. Trzciński
    Conference:
    Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, IJCAI-22 (rok: 2022, tom: International Joint Conference on Artificial Intelligence IJCAI-ECAI 2022, strony: 2902-2908), Wydawca: International Joint Conferences on Artificial Intelligence Organization
    Data:
    konferencja July 23-29, 2022
    Status:
    Published
    DOI:
    10.24963/ijcai.2022/402 - link to the publication
  15. Continual Learning of 3D Point Cloud Generators
    Authors:
    M. Sadowski, K. Piczak, P. Spurek, T. Trzciński
    Conference:
    Neural Information Processing, Part I (rok: 2021, tom: International Conference on Neural Information Processing, strony: 330-341), Wydawca: Springer International Publishing
    Data:
    konferencja 8-12 December 2021
    Status:
    Published
    DOI:
    10.1007/978-3-030-92185-9_27 - link to the publication
  16. On Analyzing Generative and Denoising Capabilities of Diffusion-based Deep Generative Models
    Authors:
    K. Deja, A. Kuzina, T. Trzciński, J. Tomczak
    Conference:
    Advances in Neural Information Processing Systems (rok: 2022, tom: Neural Information Processing Systems (NeurIPS), strony: b.d.), Wydawca: Curran Associates
    Status:
    Accepted for publication

O Narodowym Centrum Nauki

Finansowanie nauki

Współpraca zagraniczna

Centrum prasowe

Deklaracja dostępności
Przetwarzanie danych osobowych
Polityka cookies

Partnerzy