SUPER-Net: Trustworthy Image Segmentation via Uncertainty Propagation in Encoder-Decoder Networks
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2021
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915529188442112 |
|---|---|
| author | Carannante, Giuseppina Bouaynaya, Nidhal C. Dera, Dimah Fathallah-Shaykh, Hassan M. Rasool, Ghulam |
| author_facet | Carannante, Giuseppina Bouaynaya, Nidhal C. Dera, Dimah Fathallah-Shaykh, Hassan M. Rasool, Ghulam |
| contents | Deep Learning (DL) holds great promise in reshaping the industry owing to its precision, efficiency, and objectivity. However, the brittleness of DL models to noisy and out-of-distribution inputs is ailing their deployment in sensitive fields. Current models often lack uncertainty quantification, providing only point estimates. We propose SUPER-Net, a Bayesian framework for trustworthy image segmentation via uncertainty propagation. Using Taylor series approximations, SUPER-Net propagates the mean and covariance of the model's posterior distribution across nonlinear layers. It generates two outputs simultaneously: the segmented image and a pixel-wise uncertainty map, eliminating the need for expensive Monte Carlo sampling. SUPER-Net's performance is extensively evaluated on MRI and CT scans under various noisy and adversarial conditions. Results show that SUPER-Net outperforms state-of-the-art models in robustness and accuracy. The uncertainty map identifies low-confidence areas affected by noise or attacks, allowing the model to self-assess segmentation reliability, particularly when errors arise from noise or adversarial examples. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2111_05978 |
| institution | arXiv |
| publishDate | 2021 |
| record_format | arxiv |
| spellingShingle | SUPER-Net: Trustworthy Image Segmentation via Uncertainty Propagation in Encoder-Decoder Networks Carannante, Giuseppina Bouaynaya, Nidhal C. Dera, Dimah Fathallah-Shaykh, Hassan M. Rasool, Ghulam Image and Video Processing Computer Vision and Pattern Recognition Machine Learning Deep Learning (DL) holds great promise in reshaping the industry owing to its precision, efficiency, and objectivity. However, the brittleness of DL models to noisy and out-of-distribution inputs is ailing their deployment in sensitive fields. Current models often lack uncertainty quantification, providing only point estimates. We propose SUPER-Net, a Bayesian framework for trustworthy image segmentation via uncertainty propagation. Using Taylor series approximations, SUPER-Net propagates the mean and covariance of the model's posterior distribution across nonlinear layers. It generates two outputs simultaneously: the segmented image and a pixel-wise uncertainty map, eliminating the need for expensive Monte Carlo sampling. SUPER-Net's performance is extensively evaluated on MRI and CT scans under various noisy and adversarial conditions. Results show that SUPER-Net outperforms state-of-the-art models in robustness and accuracy. The uncertainty map identifies low-confidence areas affected by noise or attacks, allowing the model to self-assess segmentation reliability, particularly when errors arise from noise or adversarial examples. |
| title | SUPER-Net: Trustworthy Image Segmentation via Uncertainty Propagation in Encoder-Decoder Networks |
| topic | Image and Video Processing Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2111.05978 |