Publications by Doruk Oner | EPFL Graph Search

Enabling Uncertainty Estimation in Iterative Neural Networks

Pascal Fua, Doruk Oner, Nikita Durasov, Minh Hieu Lê

Turning pass-through network architectures into iterative ones, which use their own output as input, is a well-known approach for boosting performance. In this paper, we argue that such architectures offer an additional benefit: The convergence rate of the ...

Curran Associates2024

Topologically Better Delineation of Curvilinear Structures

Doruk Oner

Curvilinear structures are frequently observed in a variety of domains and are essential for comprehending neural circuits, detecting fractures in materials, and determining road and irrigation canal networks. It can be costly and time-consuming to manuall ...

EPFL2023

TOPO-Loss for continuity-preserving crack detection using deep learning

Pascal Fua, Radhakrishna Achanta, Katrin Beyer, Fernando Perez Cruz, Doruk Oner, Bryan German Pantoja Rosero

We present a method for segmenting cracks in images of masonry buildings damaged by earthquakes. Existing methods of crack detection fail to preserve the continuity of cracks, and their performance deteriorates with imprecise training labels. We address th ...

2022

Adjusting the Ground Truth Annotations for Connectivity-Based Learning to Delineate

Pascal Fua, Leonardo Citraro, Mateusz Kozinski, Doruk Oner

Deep learning-based approaches to delineating 3D structure depend on accurate annotations to train the networks. Yet in practice, people, no matter how conscientious, have trouble precisely delineating in 3D and on a large scale, in part because the data i ...

Institute of Electrical and Electronics Engineers2022

Persistent Homology with Improved Locality Information for more Effective Delineation

Kathryn Hess Bellwald, Pascal Fua, Mateusz Kozinski, Adélie Eliane Garin, Doruk Oner

Persistent Homology (PH) has been successfully used to train networks to detect curvilinear structures and to improve the topological quality of their results. However, existing methods are very global and ignore the location of topological features. In th ...

2022

Promoting Connectivity of Network-Like Structures by Enforcing Region Separation

Pascal Fua, Leonardo Citraro, Mateusz Kozinski, Doruk Oner

We propose a novel, connectivity-oriented loss function for training deep convolutional networks to reconstruct network-like structures, like roads and irrigation canals, from aerial images. The main idea behind our loss is to express the connectivity of r ...

2022

Enforcing Connectivity of 3D Linear Structures Using Their 2D Projections

Pascal Fua, Mateusz Kozinski, Doruk Oner, Hussein Fadl Hassan Hassan Osman

Many biological and medical tasks require the delineation of 3D curvilinear structures such as blood vessels and neurites from image volumes. This is typically done using neural networks trained by minimizing voxel-wise loss functions that do not capture t ...

SPRINGER INTERNATIONAL PUBLISHING AG2022

Drainage Canals in Southeast Asian Peatlands Increase Carbon Emissions

Mateusz Kozinski, Doruk Oner

Drainage canals associated with logging and agriculture dry out organic soils in tropical peatlands, thereby threatening the viability of long-term carbon stores due to increased emissions from decomposition, fire, and fluvial transport. In Southeast Asian ...

AMER GEOPHYSICAL UNION2021