Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 13 of 13 for “"ResNet34"”.
-
Resprop : reused sparsified backpropagation
… 8.0x and 7.2x speedup in the backward pass on ResNet34 and VGG16 versus a GTX 1080 Ti GPU.
-
Využití neoznačenačených dat pro segmentaci sítnice
… 63,4%. Použití přenosového učení pomocí kodéru ResNet34 předtrénovaného na síti ImageNet vedlo k výraznému zlepšení výkonu s koeficientem Dice 81,9%, IoU 69,3% a přesností 96,7%. Kromě toho byly ke zvýšení výkonu modelu použity strategie učení s částečným dohledem, včetně pseudoznačení a …
-
Integrating Vision–Language Models with Medical Imaging and Clinical Data for Improved Lymphoma Diagnosis
… System (LARS), an ensemble model composed of 10 ResNet34-based models trained on over 17,000 [18F] FDG-PET/CT scans. Although LARS has achieved impressive diagnostic accuracy, it relies solely on image features and lacks the capability to understand and utilize patient metadata (such as age, sex, …
-
Tuberculosis detection and localization from chest x-ray images using deep convolutional neural networks
… (CNN) models (VGG16, VGG19, Inception V3, ResNet34, ResNet50 and ResNet101) that are used for TB detection. The proposed method includes three major processes: modifications on CNN model structures, model fine-tuning via artificial bee colony algorithm, and the implementation of the …
-
Modelo de Segmentación para la identificación de caminos que deforestan la Amazonía del Perú mediante el estudio de imágenes satelitales
… learning empleando un codificador preentrenado ResNet34 como backbone dentro de la U-Net, con el fin de aprovechar características previamente aprendidas y optimizar el rendimiento en un escenario de datos limitados. Los modelos entrenados fueron evaluados mediante la métrica Intersection over …
-
Deep Learning Methods for Automatic Differentiation of Non-Ventricular Rhythms, Ventricular Tachycardia, and Ventricular Fibrillation in Electrocardiograms
… are integrated into a hierarchical multi-stream ResNet34 architecture. The proposed approach achieves an overall class sensitivity of 89.68%, with individual sensitivities of 81.46% for ventricular tachycardia, 89.29% for ventricular fibrillation, and 98.28% for non-ventricular rhythms, based on …
-
Modelling and inference of interacting cell dynamics using single time-point images
… a robust U-Net segmentation model with a ResNet34 back- bone and boundary-aware probability shaping, alongside a velocity-predictive track- ing algorithm to mitigate survival bias and extract unbiased biological features. We evaluate an agent-based model utilizing overdamped Langevin …
-
AI-Enabled Retinal Imaging for Scalable Prediction of Systemic and Ocular Diseases Using an Oculomics Framework
… VGG16 achieved 91.54% accuracy (AUC 0.9551) and ResNet34 88.78% (AUC 0.9422). Grad-CAM showed localised attention on inner and deep retinal layers in normal cases, while abnormal ICP produced diffuse patterns, especially in VGG16, consistent with clinical findings. In DR detection, DL models …
-
Quantification of crops' consistency on corn fields using robust deep learning models
… for semantic segmentation exploits the ResNet34 module into the U-Net encoder to extract features efficiently during the down-sampling process. This results in a balance of lower trainable parameters and the preservation of spatial resolution during the feature extraction process. Our …
-
Autonomous Corrosion Detection In Steel Structures Using Different Non-Contact Techniques
… backbone with DenseNet121, EfficientNetB7, and ResNet34 were tested to classify the corroded pixels. Out of all these U-shaped models, the network with ResNet34 as the backbone outperformed the other models by predicting 93.4% corroded pixels precisely with a 90.77% intersection over union (IOU) …
-
Uso de redes neurais profundas na avaliação do ataque de mosca minadora ao tomateiro
… comparado aos demais, sendo as espinhas dorsais ResNet34 e DenseNet121 as que apresentaram menor valor de RSME, o que confirma ainda os valores de IoU encontrados para estes modelos. Os experimentos computacionais demonstrados nesta pesquisa se mostraram promissores, principalmente pela …
-
Artificial intelligence in dentistry: Scoping review and bridging observed knowledge gaps via a methodological study and a clinical trial
… backbones of varying complexities (ResNet18, ResNet34, ResNet50, ResNet101, ResNet152, VGG13, VGG16, VGG19, DenseNet121, DenseNet161, DenseNet169, and Dense-Net201), and three initialization strategies (random, ImageNet, and CheXpert weights). 1,625 dental bitewing radiographs were used for …