Photo of Pavlo Melnyk

Pavlo Melnyk

PhD student

Passionate about AI in general and Geometric Deep Learning in particular: debunking the magic behind artificial neural networks by viewing and designing them from a geometry perspective.

I am an academic PhD student at Computer Vision Laboratory.

Research Activities

My PhD project, “How to Inject Geometry into Deep Learning - Theoretical Foundation and New Computational Methods,” is funded by Wallenberg AI, Autonomous Systems and Software Program (WASP).

I am part of the Geometric Deep Learning Cluster within WASP.

Teaching

I am a teacher assistant and a project supervisor in several courses such as ”TSBB06 Multi-dimensional signal analysis”, ”TSBB14 Signal- and image-processing”, and ”TSBB15 Computer vision”.

I also supervise master theses.

Publications

  • Pavlo Melnyk, Michael Felsberg, Mårten Wadenbäck. "Steerable 3D Spherical Neurons", Proceedings of the 39th International Conference on Machine Learning, PMLR 162:15330-15339, 2022.

Quick Facts: Pavlo Melnyk

CV

  • 2019-: PhD in Electrical Engineering with specialization in Computer Vision, Linköping University, Linköping, Sweden.
  • 2016-2019: Master's in Computer Science and Technology, Hunan University (湖南大学), Changsha, Hunan province, China.
  • 2012-2016: Bachelor's in Information Security Systems, Donetsk National Technical University, Pokrovsk, Ukraine.

Languages

  • Swedish (certified, C1)
  • Chinese (certified, HSK5 (2019's version))
  • English
  • Ukrainian (native/bilingual)

Publications

Pavlo Melnyk,  Spherical NeurO(n)s for Geometric Deep Learning, Linköping Studies in Science and Technology. Dissertations 2393, Linköping University Electronic Press, Linköping (2024)  https://doi.org/10.3384/9789180756808

Pavlo Melnyk, Andreas Robinson, Michael Felsberg, Mårten Wadenbäck,  TetraSphere: A Neural Descriptor for O(3)-Invariant Point Cloud Analysis, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). June 2024, Proceedings: IEEE Conference on Computer Vision and Pattern Recognition, pp. 5620-5630, IEEE Computer Society (2024)  https://doi.org/10.1109/CVPR52733.2024.00537  https://arxiv.org/abs/2211.14456

Pavlo Melnyk, Michael Felsberg, Mårten Wadenbäck, Andreas Robinson, Cuong Le,  On Learning Deep O(n)-Equivariant Hyperspheres, Proceedings of the 41st International Conference on Machine Learning, Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix (eds.), Proceedings of Machine Learning Research, pp. 35324-35339, PMLR (2024)  https://arxiv.org/abs/2305.15613

Pavlo Melnyk, Michael Felsberg, Mårten Wadenbäck,  Steerable 3D Spherical Neurons, Proceedings of the 39th International Conference on Machine Learning, Kamalika Chaudhuri, Stefanie Jegelka, Le Song, Csaba Szepesvari, Gang Niu, Sivan Sabato (eds.), Proceedings of Machine Learning Research, pp. 15330-15339, PMLR (2022)  https://arxiv.org/abs/2106.13863

Pavlo Melnyk, Michael Felsberg, Mårten Wadenbäck,  Embed Me If You Can: A Geometric Perceptron, Proceedings 2021 IEEE/CVF International Conference on Computer Vision ICCV 2021, IEEE International Conference on Computer Vision. Proceedings, pp. 1256-1264, Institute of Electrical and Electronics Engineers (IEEE) (2021)  https://doi.org/10.1109/iccv48922.2021.00131  https://arxiv.org/abs/2006.06507

Coworkers CVL

Organisation