Essays about: "Swin"
Showing result 1 - 5 of 11 essays containing the word Swin.
-
1. Evaluation of deep learning methods for industrial automation
University essay from Umeå universitet/Institutionen för datavetenskapAbstract : The rise and adaptation of the transformer architecture from natural language processing to visual tasks have proven a useful and powerful tool. Subsequent architectures such as visual transformers (ViT) and shifting window (SWIN) transformers have proven to be comparable and oftentimes exceed convolutional neural networks (CNNs) in terms of accuracy. READ MORE
-
2. Particle Detection in Bone Marrow Using CNNs and Transformer Networks
University essay from Lunds universitet/Matematik LTHAbstract : The critical role of hematological stem cells, concentrated in bone marrow, in disease diagnosis makes the detection of these cells an imperative task. The current process, however, is slow and heavily reliant on expert interpretation, underscoring the need for automation. READ MORE
-
3. Comparative Analysis of Transformer and CNN Based Models for 2D Brain Tumor Segmentation
University essay from Linköpings universitet/Institutionen för medicinsk teknikAbstract : A brain tumor is an abnormal growth of cells within the brain, which can be categorized into primary and secondary tumor types. The most common type of primary tumors in adults are gliomas, which can be further classified into high-grade gliomas (HGGs) and low-grade gliomas (LGGs). READ MORE
-
4. Regularizing Vision-Transformers Using Gumbel-Softmax Distributions on Echocardiography Data
University essay from KTH/Skolan för elektroteknik och datavetenskap (EECS)Abstract : This thesis introduces an novel approach to model regularization in Vision Transformers (ViTs), a category of deep learning models. It employs stochastic embedded feature selection within the context of echocardiography video analysis, specifically focusing on the EchoNet-Dynamic dataset. READ MORE
-
5. Self-supervised pre-training of an attention-based model for 3D medical image segmentation
University essay from KTH/Skolan för elektroteknik och datavetenskap (EECS)Abstract : Accurate segmentation of anatomical structures is crucial for radiation therapy in cancer treatment. Deep learning methods have been demonstrated effective for segmentation of 3D medical images, establishing the current standard. However, they require large amounts of labelled data and suffer from reduced performance on domain shift. READ MORE