Cross-modal generalization performance of FocusGate-Net.

Cross-modal generalization performance of FocusGate-Net.

<p>Cross-modal generalization performance of FocusGate-Net.</p>

محفوظ في:

التفاصيل البيبلوغرافية
المؤلف الرئيسي:	Şafak Kılıç (22227019) (author)
منشور في:	2025
الموضوعات:	Biophysics Medicine Biotechnology Space Science Biological Sciences not elsewhere classified Information Systems not elsewhere classified thereby enhancing applicability suboptimal boundary localization preserving low latency models often struggle grained spatial context enhancing feature selectivity dataset generalization capability challenging task requiring skin lesion segmentation convolution hybrid network achieving dice scores minimal architectural complexity time clinical settings paper introduces focusgate convolutional feature extractors balance segmentation accuracy based hybrid architectures attention guided mlp clinical settings polyp segmentation mlp architectures hybrid design hybrid architecture dice coefficient based modules based methods attention gate architectural simplicity architectural component unet ++, results highlight promising alternative lightweight mlp inference speed frequently lack evaluate focusgate design aims dermatoscopic images computational efficiency capture fine approach leverages
الوسوم:	إضافة وسم لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!

مواد مشابهة

FocusGate-Net architecture overview.
حسب: Şafak Kılıç (22227019)
منشور في: (2025)

Generalization results of FocusGate-Net on PH2 and Kvasir-SEG.
حسب: Şafak Kılıç (22227019)
منشور في: (2025)

Attention Gate (AG) architecture for skip connections.
حسب: Şafak Kılıç (22227019)
منشور في: (2025)

Comparison with state-of-the-art Transformer and MLP-based models on ISIC2018.
حسب: Şafak Kılıç (22227019)
منشور في: (2025)

Benchmark comparison of segmentation models on ISIC2018.
حسب: Şafak Kılıç (22227019)
منشور في: (2025)

Ablation study results on ISIC2018.
حسب: Şafak Kılıç (22227019)
منشور في: (2025)

Heatmap of evaluation metrics across models on the ISIC2018 dataset.
حسب: Şafak Kılıç (22227019)
منشور في: (2025)

Qualitative comparison of segmentation outputs on ISIC2018.
حسب: Şafak Kılıç (22227019)
منشور في: (2025)

Dice and IoU Comparison Across Models on ISIC2018.
حسب: Şafak Kılıç (22227019)
منشور في: (2025)

Shifted Token MLP (ST-MLP) module architecture.
حسب: Şafak Kılıç (22227019)
منشور في: (2025)

Testing set assessment metrics for prediction of surface coal mine valley fill faces using different architectural configurations and varying sample sizes.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Testing set assessment metrics for prediction of historic mine benches using different architectural configurations and varying sample sizes.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Testing set assessment metrics for prediction of agricultural terraces using different architectural configurations and varying sample sizes.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Assessment metrics used in study. TP = true positive; TN = true negative; FP = false positive; FN = false negative.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Model complexity and computational cost comparison. 1 GFLOP = 1 billion FLOPs; 1 GMAC = 1 billion MACs.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Summary of UNet-based models with descriptions of each configuration.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Overview of the available training, validation, and test chips for each dataset used.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Bootstrap analysis of overall accuracy, F1-score, precision, and recall across model configurations for mineBenchDL, terraceDL and vfillDL datasets.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Comparison of valley fill faces detection results using different architectural configurations.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Comparison of historic mine benches detection results using different architectural configurations.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Comparison of agricultural terraces detection results using different architectural configurations.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

F1-score calculated from the withheld test data using different architectural configurations and training set sizes.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Validation F1-score for terraceDL, mineBenchDL, and vfillDL datasets using varying training sample sizes and different model configurations across 25 training epochs.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Training loss for terraceDL, mineBenchDL, and vfillDL datasets using varying training sample sizes and different model configurations across 25 training epochs.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Training loss for terraceDL, mineBenchDL, and vfillDL datasets using all training samples and different model configurations across 25 training epochs.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Training, test, and validation areas for (a) agricultural terraces (terraceDL) dataset [29] in Iowa, USA; (b) surface coal mining valley fill faces (vfillDL) dataset [30] in southern West Virginia, eastern Kentucky, and southwestern Virginia, USA; and (c) historic mine benches (minebenchDL) dataset [31] in northern West Virginia, USA.
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

UNet implementation provided by geodl [27].
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

AG module mechanism [61].
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Squeeze and excitation (SE) module [58].
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Dilated convolution (DC) module implemented in geodl [27].
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Residual connection [2] within a double-convolution block using an identity connection (a) and projection connection (b).
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

UNet architecture [1].
حسب: Sarah Farhadpour (20390793)
منشور في: (2025)

Comparison of segmentation IoU values for RETINA versus benchmark models, including: randomly initialized (Rand. Init.) UNet-ResNet50, UNet-ResNet50 pre-trained on CEM500K, Rand. Init. 2D TransUNet, Rand. Init. 3D TransUNet and Rand. Init. nnUNet.
حسب: Cheng Xing (723209)
منشور في: (2025)

Visual comparison of segmentation.
حسب: Cheng Xing (723209)
منشور في: (2025)

RETINA demonstrates fast convergence and robustness, outperforming other models across all segmentation benchmarks.
حسب: Cheng Xing (723209)
منشور في: (2025)

Augmented CEM500K images and benchmark datasets.
حسب: Cheng Xing (723209)
منشور في: (2025)

Overview of the RETINA method.
حسب: Cheng Xing (723209)
منشور في: (2025)

3D visual comparison of segmentation.
حسب: Cheng Xing (723209)
منشور في: (2025)

Analysis of failure segmentation results from LKDA-Net and other methods on Synapse and BTCV datasets. These segmentation results are appraised based on the average Dice score for each case.
حسب: Ming Li (91180)
منشور في: (2025)

The ablation experiment data of the LKDA-Net model on the Synapse dataset. We exhibit the parameter number and model performance when our model is incorporated into the baseline, and present the performance of LKD Attention with different convolution kernel sizes.
حسب: Ming Li (91180)
منشور في: (2025)

Cannot write session to /tmp/vufind_sessions/sess_r3g2s2lnlolmltquhors7stovd