Publications | ITU Vision Lab

2024

ICLR

Calibrating Bayesian UNet++ for Sub-Seasonal Forecasting

B. Asan , A. Akgül , A. Unal , and 2 more authors

In Tackling Climate Change with Machine Learning at ICLR 2024 , 2024

@inproceedings{asan2024calibrating,
  title = {Calibrating Bayesian UNet++ for Sub-Seasonal Forecasting},
  author = {Asan, B. and Akgül, A. and Unal, A. and Kandemir, M. and Unal, G.},
  booktitle = {Tackling Climate Change with Machine Learning at ICLR 2024},
  year = {2024},
  url = {https://arxiv.org/abs/2403.16612},
}

L4DC

Continual Learning of Multi-modal Dynamics with External Memory

A. Akgül , G. Unal , and M. Kandemir

In Proceedings of The 6th Annual Learning for Dynamics and Control Conference , 2024

Abs Bib HTML

@inproceedings{akgul2024cddp,
  title = {Continual Learning of Multi-modal Dynamics with External Memory},
  author = {Akgül, A. and Unal, G. and Kandemir, M.},
  booktitle = {Proceedings of The 6th Annual Learning for Dynamics and Control Conference},
  year = {2024},
  url = {https://arxiv.org/abs/2203.00936},
}

IEEE FG

epsilon-Mesh Attack: A Surface-based Adversarial Point Cloud Attack for Facial Expression Recognition

Batuhan Cengiz , Mert Gulsen , Yusuf H Sahin , and 1 more author

2024

Abs Bib HTML

@article{cengiz2024epsilon,
  title = {epsilon-Mesh Attack: A Surface-based Adversarial Point Cloud Attack for Facial Expression Recognition},
  author = {Cengiz, Batuhan and Gulsen, Mert and Sahin, Yusuf H and Unal, Gozde},
  booktitle = {IEEE International Conference on Automatic Face and Gesture Recognition},
  year = {2024},
  url = {https://arxiv.org/abs/2403.06661v1},
}

ACML

Protodiffusion: Classifier-free diffusion guidance with prototype learning

Gulcin Baykal , Halil Faruk Karagoz , Taha Binhuraib , and 1 more author

In Asian Conference on Machine Learning , 2024

Abs Bib HTML

@inproceedings{baykal2024protodiffusion,
  title = {Protodiffusion: Classifier-free diffusion guidance with prototype learning},
  author = {Baykal, Gulcin and Karagoz, Halil Faruk and Binhuraib, Taha and Unal, Gozde},
  booktitle = {Asian Conference on Machine Learning},
  year = {2024},
  url = {https://proceedings.mlr.press/v222/baykal24a.html},
}

2023

TMLR

Meta Continual Learning on Graphs with Experience Replay

A. Unal , A. Akgül , M. Kandemir , and 1 more author

2023

Abs Bib HTML

@article{unal2023meta,
  title = {Meta Continual Learning on Graphs with Experience Replay},
  author = {Unal, A. and Akg{\"u}l, A. and Kandemir, M. and Unal, G.},
  booktitle = {Transactions on Machine Learning Research},
  year = {2023},
  url = {https://openreview.net/forum?id=8tnrh56P5W},
}

arXiv

Textile Pattern Generation Using Diffusion Models

Halil Faruk Karagoz , Gulcin Baykal , Irem Arikan Eksi , and 1 more author

2023

Abs Bib HTML

@article{karagoz2023textile,
  title = {Textile Pattern Generation Using Diffusion Models},
  author = {Karagoz, Halil Faruk and Baykal, Gulcin and Eksi, Irem Arikan and Unal, Gozde},
  booktitle = {arXiv Preprint},
  year = {2023},
  url = {https://arxiv.org/abs/2304.00520},
}

arXiv
EdVAE: Mitigating Codebook Collapse with Evidential Discrete Variational Autoencoders

Gulcin Baykal , Melih Kandemir , and Gozde Unal

2023

Abs Bib HTML

Codebook collapse is a common problem in training deep generative models with discrete representation spaces like Vector Quantized Variational Autoencoders (VQ-VAEs). We observe that the same problem arises for the alternatively designed discrete variational autoencoders (dVAEs) whose encoder directly learns a distribution over the codebook embeddings to represent the data. We hypothesize that using the softmax function to obtain a probability distribution causes the codebook collapse by assigning overconfident probabilities to the best matching codebook elements. In this paper, we propose a novel way to incorporate evidential deep learning (EDL) instead of softmax to combat the codebook collapse problem of dVAE. We evidentially monitor the significance of attaining the probability distribution over the codebook embeddings, in contrast to softmax usage. Our experiments using various datasets show that our model, called EdVAE, mitigates codebook collapse while improving the reconstruction performance, and enhances the codebook usage compared to dVAE and VQ-VAE based models.
@article{baykal2023edvae, title = {EdVAE: Mitigating Codebook Collapse with Evidential Discrete Variational Autoencoders}, author = {Baykal, Gulcin and Kandemir, Melih and Unal, Gozde}, booktitle = {arXiv Preprint}, year = {2023}, url = {https://arxiv.org/abs/2310.05718v2}, }

MVA

TinyPedSeg: A Tiny Pedestrian Segmentation Benchmark for Top-Down Drone Images

Yusuf H Sahin , Elvin Abdinli , M Arda Aydın , and 1 more author

In International Conference on Machine Vision Applications , 2023

Abs Bib HTML

@inproceedings{sahin2023tinypedseg,
  title = {TinyPedSeg: A Tiny Pedestrian Segmentation Benchmark for Top-Down Drone Images},
  author = {Sahin, Yusuf H and Abdinli, Elvin and Ayd{\i}n, M Arda and Unal, Gozde},
  booktitle = {International Conference on Machine Vision Applications},
  year = {2023},
  url = {https://ieeexplore.ieee.org/abstract/document/10215829},
}

MDPI

ALReg: Registration of 3D Point Clouds Using Active Learning

Yusuf Huseyin Sahin , Oguzhan Karabacak , Melih Kandemir , and 1 more author

2023

Abs Bib HTML

@article{sahin2023alreg,
  title = {ALReg: Registration of 3D Point Clouds Using Active Learning},
  author = {Sahin, Yusuf Huseyin and Karabacak, Oguzhan and Kandemir, Melih and Unal, Gozde},
  booktitle = {MDPI Applied Sciences},
  year = {2023},
  url = {https://www.mdpi.com/2076-3417/13/13/7422},
}

arXiv

GaussianMLR: Learning Implicit Class Significance via Calibrated Multi-Label Ranking

V Bugra Yesilkaynak , Emine Dari , Alican Mertan , and 1 more author

2023

Abs Bib HTML

@article{yesilkaynak2023gaussianmlr,
  title = {GaussianMLR: Learning Implicit Class Significance via Calibrated Multi-Label Ranking},
  author = {Yesilkaynak, V Bugra and Dari, Emine and Mertan, Alican and Unal, Gozde},
  booktitle = {arXiv Preprint},
  year = {2023},
  url = {https://arxiv.org/abs/2303.03907},
}

2022

NeuRIPS

How to combine variational bayesian networks in federated learning

Atahan Ozer , Kadir Burak Buldu , Abdullah Akgül , and 1 more author

2022

Abs Bib HTML

@article{ozer2022combine,
  title = {How to combine variational bayesian networks in federated learning},
  author = {Ozer, Atahan and Buldu, Kadir Burak and Akg{\"u}l, Abdullah and Unal, Gozde},
  booktitle = {International Workshop on Federated Learning: Recent Advances and New Challenges in Conjunction with NeurIPS 2022 (FL-NeurIPS'22)},
  year = {2022},
  url = {https://arxiv.org/abs/2206.10897},
}

ICPRAI

UGQE: Uncertainty Guided Query Expansion

Firat Oncel , Mehmet Aygün , Gulcin Baykal , and 1 more author

In International Conference on Pattern Recognition and Artificial Intelligence , 2022

Abs Bib HTML

@inproceedings{oncel2022ugqe,
  title = {UGQE: Uncertainty Guided Query Expansion},
  author = {Oncel, Firat and Ayg{\"u}n, Mehmet and Baykal, Gulcin and Unal, Gozde},
  booktitle = {International Conference on Pattern Recognition and Artificial Intelligence},
  year = {2022},
  url = {https://link.springer.com/chapter/10.1007/978-3-031-09037-0_10},
}

ODFNet: Using orientation distribution functions to characterize 3D point clouds

Yusuf H Sahin , Alican Mertan , and Gozde Unal

2022

Abs Bib HTML

@article{sahin2022odfnet,
  title = {ODFNet: Using orientation distribution functions to characterize 3D point clouds},
  author = {Sahin, Yusuf H and Mertan, Alican and Unal, Gozde},
  booktitle = {Computer and Graphics},
  year = {2022},
  url = {https://www.sciencedirect.com/science/article/pii/S0097849321001801},
}

DSP

Single image depth estimation: An overview

Alican Mertan , Damien Jade Duff , and Gozde Unal

2022

Abs Bib HTML

@article{mertan2022single,
  title = {Single image depth estimation: An overview},
  author = {Mertan, Alican and Duff, Damien Jade and Unal, Gozde},
  booktitle = {Digital Signal Processing},
  year = {2022},
  url = {https://www.sciencedirect.com/science/article/pii/S1051200422000586},
}

Exploring deshufflegans in self-supervised generative adversarial networks

Gulcin Baykal , Furkan Ozcelik , and Gozde Unal

2022

Abs Bib HTML

@article{baykal2022exploring,
  title = {Exploring deshufflegans in self-supervised generative adversarial networks},
  author = {Baykal, Gulcin and Ozcelik, Furkan and Unal, Gozde},
  booktitle = {Pattern Recognition},
  year = {2022},
  url = {https://www.sciencedirect.com/science/article/pii/S0031320321004167},
}

ICLR

Evidential turing processes

Melih Kandemir , Abdullah Akgül , Manuel Haussmann , and 1 more author

2022

Abs Bib HTML

@article{kandemir2021evidential,
  title = {Evidential turing processes},
  author = {Kandemir, Melih and Akg{\"u}l, Abdullah and Haussmann, Manuel and Unal, Gozde},
  booktitle = {International Conference on Learning Representations},
  year = {2022},
  url = {https://arxiv.org/abs/2106.01216},
}

2021

PRIME

Uncertainty-Based Dynamic Graph Neighborhoods for Medical Segmentation

Ufuk Demir , Atahan Ozer , Yusuf H Sahin , and 1 more author

In International Workshop on PRedictive Intelligence In MEdicine , 2021

Abs Bib HTML

@inproceedings{demir2021uncertainty,
  title = {Uncertainty-Based Dynamic Graph Neighborhoods for Medical Segmentation},
  author = {Demir, Ufuk and Ozer, Atahan and Sahin, Yusuf H and Unal, Gozde},
  booktitle = {International Workshop on PRedictive Intelligence In MEdicine},
  year = {2021},
  url = {https://link.springer.com/chapter/10.1007/978-3-030-87602-9_24},
}