Zobrazeno 1 - 10
of 21 449
pro vyhledávání: '"skip-connection"'
Autor:
Zhang, Qiong1,2 (AUTHOR) 18862928127@163.com, Min, Byungwon2 (AUTHOR), Hang, Yiliu1 (AUTHOR), Chen, Hao1,2 (AUTHOR), Qiu, Jianlin1 (AUTHOR)
Publikováno v:
Scientific Reports. 10/5/2024, Vol. 14 Issue 1, p1-13. 13p.
Training deep Convolutional Neural Networks (CNNs) presents unique challenges, including the pervasive issue of elimination singularities, consistent deactivation of nodes leading to degenerate manifolds within the loss landscape. These singularities
Externí odkaz:
http://arxiv.org/abs/2409.13154
Publikováno v:
Scientific Reports, Vol 14, Iss 1, Pp 1-13 (2024)
Abstract The segmentation accuracy of the lung images is affected by the occlusion of the front background objects. To address this problem, we propose a full-scale lung image segmentation algorithm based on hybrid skip connection and attention mecha
Externí odkaz:
https://doaj.org/article/c6d9dfc25891400fa38741e1ab007e1b
Autor:
Al Mokhtar, Zahraa T.1 zahraatalal84@gmail.com, Dawwd, Shefa A.1
Publikováno v:
Iraqi Journal for Electrical & Electronic Engineering. Jun2024, Vol. 20 Issue 1, p195-205. 11p.
Deep learning has made significant progress in computer vision, specifically in image classification, object detection, and semantic segmentation. The skip connection has played an essential role in the architecture of deep neural networks,enabling e
Externí odkaz:
http://arxiv.org/abs/2405.01725
Autor:
Zhang, Qi1 (AUTHOR), Geng, Guohua1 (AUTHOR) ghgeng@nwu.edu.cn, Zhou, Pengbo2 (AUTHOR) zhoupengbo@bnu.edu.cn, Liu, Qinglin1 (AUTHOR), Wang, Yong1 (AUTHOR), Li, Kang1 (AUTHOR)
Publikováno v:
Remote Sensing. Oct2024, Vol. 16 Issue 19, p3622. 22p.
Breast cancer is a major global health concern. Pathologists face challenges in analyzing complex features from pathological images, which is a time-consuming and labor-intensive task. Therefore, efficient computer-based diagnostic tools are needed f
Externí odkaz:
http://arxiv.org/abs/2406.08604
Ultrasound is a widely used medical tool for non-invasive diagnosis, but its images often contain speckle noise which can lower their resolution and contrast-to-noise ratio. This can make it more difficult to extract, recognize, and analyze features
Externí odkaz:
http://arxiv.org/abs/2403.02750
We tested front-end enhanced neural models where a frozen classifier was prepended by a differentiable and fully convolutional model with a skip connection. By training them using a small learning rate for about one epoch, we obtained models that ret
Externí odkaz:
http://arxiv.org/abs/2402.17018
Autor:
Ataiefard, Foozhan, Ahmed, Walid, Hajimolahoseini, Habib, Asani, Saina, Javadi, Farnoosh, Hassanpour, Mohammad, Awad, Omar Mohamed, Wen, Austin, Liu, Kangling, Liu, Yang
Vision transformers are known to be more computationally and data-intensive than CNN models. These transformer models such as ViT, require all the input image tokens to learn the relationship among them. However, many of these tokens are not informat
Externí odkaz:
http://arxiv.org/abs/2401.15293