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 20 of 78 for “"DenseNet"”.

  1. 3D Densely Connected Convolutional Network for the Recognition of Human Shopping Actions

    … a 3D Densely Connected Convolutional Network (3D-DenseNet) for action recognition that can have more than 100 layers without exhibiting performance degradation or overfitting. Our network expands Densely Connected Convolutional Networks (DenseNet) [32] to 3D-DenseNet by adding the temporal …

    ottawa-retro Repository record for 3D Densely Connected Convolutional Network for the Recognition of Human Shopping Actions (opens in a new tab)

  2. Facial Action Unit Detection With Deep Convolutional Neural Networks

    … each layer of the network. We introduce Modified DenseNet considering DenseNet as a baseline model. Averaging all the features obtained from each block of DenseNet gives importance to each level of features which can get lost during concatenating the layers in DenseNet and other state of the art …

    denver Repository record for Facial Action Unit Detection With Deep Convolutional Neural Networks (opens in a new tab)

  3. OSTEOPOROSIS DISEASE DETECTION USING HYBRID DEEP LEARNING

    … the architectural strengths of VGG19 and DenseNet-121 for the enhanced detection and classification of osteoporosis from X-ray images. The comprehensive utilization of the deep feature extraction capabilities of VGG19, with its sequential convolutional framework, and the …

    csusb Repository record for OSTEOPOROSIS DISEASE DETECTION USING HYBRID DEEP LEARNING (opens in a new tab)

  4. Avaliação de desempenho preditivo de redes neurais em aprendizagem federada para classificação de estágios da doença de Alzheimer

    … ideal. Avaliamos duas arquiteturas (ResNet-50 e DenseNet-169) em cenários federados com 3, 5 e 7 clientes, utilizando três algoritmos de agregação: FedAvg, FedAvgM e FedProx. Os experimentos foram conduzidos em duas fases: Fase I (548 pacientes) e Fase II (1.287 pacientes). Na Fase I, …

    brazil-ufrgs Repository record for Avaliação de desempenho preditivo de redes neurais em aprendizagem federada para classificação de estágios da doença de Alzheimer (opens in a new tab)

  5. Classification of Volume Regions as Cancer Tumors in Three-Dimensional Breast Magnetic Resonance Imaging Using Deep Learning and Convolutional Neural Networks

    … CNN, τα Πυκνά Συνεδεδεμένα Συνελικτικά Δίκτυα (DenseNet) έχουν κερδίσει ιδιαίτερη προσοχή λόγω της καλής απόδοσης τους σε τέτοιου τύπου εργασίες. Αυτή η διπλωματική εργασία επικεντρώνεται στην εφαρμογή αρχιτεκτονικών DenseNet στην ταξινόμηση περιοχών όγκου ως καρκινικούς όγκους σε τρισδιάστατες …

    athens Repository record for Classification of Volume Regions as Cancer Tumors in Three-Dimensional Breast Magnetic Resonance Imaging Using Deep Learning and Convolutional Neural Networks (opens in a new tab)

  6. An Efficient Neural Network Architecture and Training Protocol for 3D Point Cloud Classification

    … points from the whole point cloud. Whereas the DenseNet is a lightweight network that takes a large number of points as input. SparseNet is composed of more number of channels than DenseNet making it more computationally expensive than DenseNet. While the accuracy of the model shows good …

    concordia Download PDF for An Efficient Neural Network Architecture and Training Protocol for 3D Point Cloud Classification (opens in a new tab)

  7. KimiaNet: Training a Deep Network for Histopathology using High-Cellularity

    … This model is trained based on the existing DenseNet-121 architecture but by using more than 240,000 image patches of 1000 ⨉ 1000 pixels acquired at 20⨉ magnification. Considering the high cost of histopathology image annotation, which makes the idea impractical at a large scale, a …

    waterloo Repository record for KimiaNet: Training a Deep Network for Histopathology using High-Cellularity (opens in a new tab)

  8. Investigation of generalization properties of convolutional neural networks /

    … neural network architecture (ResNet and DenseNet with different number of layers) on convolutional neural network training and generalization and compared with ensembled network models. Investigated the impact of data augmentation on image level (CIFAR-10 and CIFAR-100 datasets), …

    vilnius Repository record for Investigation of generalization properties of convolutional neural networks / (opens in a new tab)

  9. ENHANCING ACCIDENT INVESTIGATION USING TRAFFIC CCTV FOOTAGE

    … Experience Project investigated how the densenet-161 model will perform on accident severity prediction compared to proposed methods. The research questions are: (Q1) What is the impact of usage of augmentation techniques on imbalanced datasets? (Q2) How will the hyper parameter tuning …

    csusb Repository record for ENHANCING ACCIDENT INVESTIGATION USING TRAFFIC CCTV FOOTAGE (opens in a new tab)

  10. Advanced deep learning towards improving prediction outcomes on medical imaging data and radiology reports

    … Classification, are used, applying these DenseNet-121 and EfficientNet architectures. In the context of our research, the Feature extraction for Content-based image retrieval with kNN approach demonstrated significant potential, with the highest F1 scores achieved as follows: DenseNet-121 …

    essex Download PDF for Advanced deep learning towards improving prediction outcomes on medical imaging data and radiology reports (opens in a new tab)

  11. Use of anchors to improve gravity-alignment of panoramas

    … the panorama, particularly near the poles. Using DenseNet-121 and ConvNeXt-tiny backbones pre-trained on ImageNet, AnchorsUp achieves state-of-the-art results on the SUN360 dataset, with 99.5% accuracy within a 5° error tolerance and 99.8% within 12°. We noted that the SWHDC technique improved the …

    brazil-ufrgs Repository record for Use of anchors to improve gravity-alignment of panoramas (opens in a new tab)

  12. Assessment of CNN+XGBoost Performance for Image Classification

    … — AlexNet, AllConvolutionalNet, WideResNet, DenseNet and CaffeNet (in transfer learning mode) — to extract features from the CNN part with different quality, which is defined by the classification accuracy of the appropriate CNN model. Then XGBoost is trained on the extracted features and the …

    mcmaster Repository record for Assessment of CNN+XGBoost Performance for Image Classification (opens in a new tab)

  13. Advancing Hyperspectral Wetland Classification via Convolutional Vision Transformer, Deformable Multiscale Network, and Layer-wise Closed-Form Learning

    … and vision transformers. ConViT integrates DenseNet-based spatial–spectral feature extraction with a novel convolutional–transformer block that preserves spatial structure, employs learnable local embeddings, and utilizes skip-connected encoders to capture both local details and long-range …

    queens Repository record for Advancing Hyperspectral Wetland Classification via Convolutional Vision Transformer, Deformable Multiscale Network, and Layer-wise Closed-Form Learning (opens in a new tab)

  14. Pigs' BCS Estimation using Computer Vision and Deep Learning Approaches

    … deep learning pipeline combining ResNet-50 and DenseNet-121 architectures with integrated depth information from Intel RealSense D435 cameras. The system was trained and validated on a dataset of 268 pigs across six pens collected from Virginia State University farms, with video streams captured …

    vt Repository record for Pigs' BCS Estimation using Computer Vision and Deep Learning Approaches (opens in a new tab)

  15. CTP in Posterior circulation infarction

    … expert clinician panel versus a deep learning, Densenet model, in POCI. The Densenet model was superior to the expert clinician panel. The degree of improvement by the Densenet model varied by individual clinician. We proposed that a hybrid clinician-densenet approach may standardise clinician …

    unsw Repository record for CTP in Posterior circulation infarction (opens in a new tab)

  16. Využití sítí typu GAN pro zpřesňování detekce a rozpoznávání dopravních značek

    … a Wasserstein GAN, kombinující architektury DenseNet a U-Net. Modely byly navrženy k syntéze snímků reálně vypadajících dopravních značek z obrázků jejich norem. Model pro detekci objektů typu SSD, natrénován pouze na syntetických datech, dosáhl přesnosti 59.6 % mAP, což je o 9.4 % lepší …

    brno-tech Repository record for Využití sítí typu GAN pro zpřesňování detekce a rozpoznávání dopravních značek (opens in a new tab)

  17. Deep Learning Optimization and Acceleration

    … networks (e.g., LeNet-5, VGG-16, GoogLeNet, DenseNet, ResNet). Experimental results are promising when compared to other state-of-the-art deep neural network acceleration efforts in the literature.

    unt Repository record for Deep Learning Optimization and Acceleration (opens in a new tab)

  18. A deep learning approach to predict crashworthiness behaviour of mechanical meta-material

    … densely connected convolutional neural network (DenseNet) were used to predict the crashworthiness behaviour of optimum mechanical meta-material structures. LMB and DenseNet were also used to predict optimum meta-material structure when peak crush force was given as input for the network. It was …

    qu-belfast Repository record for A deep learning approach to predict crashworthiness behaviour of mechanical meta-material (opens in a new tab) Download PDF for A deep learning approach to predict crashworthiness behaviour of mechanical meta-material (opens in a new tab)

  19. Improving classification performance for the Functional Map of the World Using Siamese and LiteDenseNet Architectures

    … deep convolutional neural networks as resnet50, densenet161, efficientnetb4, and regnet_y_1_6gf. Our results show that the developed Siamese architectures learn their internal representations differently from their regular counterparts. Siamese representations are wider, richer, more …

    mcmaster Repository record for Improving classification performance for the Functional Map of the World Using Siamese and LiteDenseNet Architectures (opens in a new tab)

  20. DropBack : continuous pruning during deep neural network training

    … on modern, compact network architectures such as Densenet and WRN-28-10, DropBack outperforms the current state-of-the- art pruning techniques in both accuracy and off-chip memory storage required for weights. On the CIFAR-10 dataset, we observe 5× reduction in weights on an already 9×-reduced …

    ubc Repository record for DropBack : continuous pruning during deep neural network training (opens in a new tab) Download PDF for DropBack : continuous pruning during deep neural network training (opens in a new tab)

Page 1 of 4