[HTML][HTML] Large-scale road extraction from high-resolution remote sensing images based on a weakly-supervised structural and orientational consistency constraint …

M Zhou, H Sui, S Chen, J Liu, W Shi, X Chen - ISPRS Journal of …, 2022‏ - Elsevier
Fully supervised road segmentation neural networks from remote sensing images rely on a
large number of densely labeled road samples, limiting their potential in large-scale …

Test time adaptation with regularized loss for weakly supervised salient object detection

O Veksler - Proceedings of the IEEE/CVF Conference on …, 2023‏ - openaccess.thecvf.com
It is well known that CNNs tend to overfit to the training data. Test-time adaptation is an
extreme approach to deal with overfitting: given a test image, the aim is to adapt the trained …

Single stage weakly supervised semantic segmentation of complex scenes

P Akiva, K Dana - Proceedings of the IEEE/CVF Winter …, 2023‏ - openaccess.thecvf.com
The costly process of obtaining semantic segmentation labels has driven research towards
weakly supervised semantic segmentation (WSSS) methods, using only image-level, point …

Advancing spatial map** for satellite image road segmentation with multi-head attention

K Ben Salah, M Othmani, J Fourati, M Kherallah - The Visual Computer, 2024‏ - Springer
Remote sensing imaging is an interesting field, particularly in road areas. Road
segmentation has become crucial in several areas, such as transportation network …

Unsupervised Losses for Clustering and Segmentation of Images: Theories & Optimization Algorithms

Z Zhang - 2024‏ - uwspace.uwaterloo.ca
Unsupervised losses are common for tasks with limited human annotations. In clustering,
they are used to group data without any labels. In semi-supervised or weakly-supervised …

Higher-order Losses and Optimization for Low-level and Deep Segmentation

D Marin - 2021‏ - uwspace.uwaterloo.ca
Regularized objectives are common in low-level and deep segmentation. Regularization
incorporates prior knowledge into objectives or losses. It represents constraints necessary to …