DMFC-UFormer: Depthwise multi-scale factorized convolution transformer-based UNet for medical image segmentation DOI
Anass Garbaz, Yassine Oukdach, Said Charfi

et al.

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 101, P. 107200 - 107200

Published: Nov. 26, 2024

Language: Английский

Combined deep convolutional neural networks for abnormality classification in wireless capsule endoscopy images DOI
Anass Garbaz, Samira Lafraxo, Said Charfi

et al.

Multimedia Tools and Applications, Journal Year: 2025, Volume and Issue: unknown

Published: March 26, 2025

Language: Английский

Citations

0

MLFA-UNet: A Multi-Level Feature Assembly UNet for Medical Image Segmentation DOI
Anass Garbaz, Yassine Oukdach, Said Charfi

et al.

Methods, Journal Year: 2024, Volume and Issue: 232, P. 52 - 64

Published: Oct. 30, 2024

Language: Английский

Citations

3

Early stopping strategies in Deep Image Prior DOI Creative Commons
Alessandro Benfenati, Ambra Catozzi, Giorgia Franchini

et al.

Soft Computing, Journal Year: 2025, Volume and Issue: unknown

Published: May 17, 2025

Language: Английский

Citations

0

Early stopping strategies in Deep Image Prior DOI Creative Commons
Alessandro Benfenati, Ambra Catozzi, Giorgia Franchini

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: May 30, 2024

Abstract The so-called Deep Image Prior approach is an unsupervised deep learning methodology which has gained great interest in recent years due to its effectiveness tackling imaging problems. However, a well known drawback of the need for proper early stopping technique prevent undesired corruptions reconstructed images. Determining optimal number iterations depends on both specific application and features images being processed. As consequence, several numerical trials are typically required decide when stop procedure, resulting significant computational costs time requirements. This paper aims introduce two techniques Prior, based different approaches perspectives. first one relies Neural Architecture Search strategy by generating hyperparameters configurations neural network employed that shall be able provide clean comparable those obtained standard configuration, optimally stopped, but with significantly fewer iterations. second proposed modified version BRISQUE metric, no-reference image quality measure, it track behaviour PSNR curve, applying without knowing ground truth image. While NAS-based particularly suited situations where limited, this latter also relevant larger allowed. Several experiments denoising applications show promising performance combined suggested procedures.

Language: Английский

Citations

0

A Systematic Review of Ulcer Detection Methods in Wireless Capsule Endoscopy DOI Creative Commons
Ahmmad Musha, Rehnuma Hasnat, Abdullah Al Mamun

et al.

Informatics in Medicine Unlocked, Journal Year: 2024, Volume and Issue: 51, P. 101600 - 101600

Published: Jan. 1, 2024

Language: Английский

Citations

0

DMFC-UFormer: Depthwise multi-scale factorized convolution transformer-based UNet for medical image segmentation DOI
Anass Garbaz, Yassine Oukdach, Said Charfi

et al.

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 101, P. 107200 - 107200

Published: Nov. 26, 2024

Language: Английский

Citations

0