Convolutional Neural Networks

Preferred Label : Convolutional Neural Networks;

MeSH definition : Artificial neural networks designed to automatically learn features through the use of sequence of layers which transform one volume to the next. Convolutional neural networks initially produce a feature map by use of convolution layers consisting of a set of learnable filters convolved with input, i.e., having smaller widths and heights and the same depth as that of input volume.;

Définition CISMeF : A specialized deep neural network architecture that consists of one or more convolution layers that is suited for processing grid-like data, such as images. In a convolution layer, a “filter” (window or template) slides over regions of the input image to identify low-level patterns (e.g., edges) by applying convolution (a mathematical dot operation applied to the input data). Different filters can be applied to extract different features, such as edges, textures, or curves in images. Additionally, CNNs can include pooling layers, whose function is to reduce the feature dimensionality while retaining relevant features. These convolution and pooling layers get stacked on top of each other to enable this network to build up a hierarchical understanding of patterns and makes CNNs effective at tasks like image recognition and computer vision. An important aspect of this network is its ability to conserve spatial information of the original input while still performing the feature extraction; Source: Adapted from: Gu, J., Wang, Z., Kuen, J., Ma, L., Shahroudy, A., Shuai, B., Liu, T., Wang, X., Wang, G., Cai, J., & Chen, T. (2018). Recent advances in convolutional neural networks. Pattern Recognition, 77, 354–377. https://doi.org/10.1016/j.patcog.2017.10.013External Link Disclaimer International Organization for Standardization. (2022). Information technology — Artificial intelligence — Artificial intelligence concepts and terminology (ISO/IEC 22989:2022). https://www.iso.org/standard/74296.htmlExternal Link Disclaimer Li, Z., Liu, F., Yang, W., Peng, S., & Zhou, J. (2022). A Survey of Convolutional Neural Networks: Analysis, Applications, and Prospects. IEEE Transactions on Neural Networks and Learning Systems, 33(12), 6999–7019. https://doi.org/10.1109/tnnls.2021.3084827;

MeSH synonym : Convolutional Neural Network; Neural Network, Convolutional;

CISMeF acronym : ConvNet; CNN;

MeSH hyponym : LeNet 5 Convolutional Neural Network; LeNet; VGGNets; Visual Geometry Group Network; ZFNets; GoogLeNets; Neural Network, Residual; Residual Neural Networks; ResNets; DenseNet;

Details

Origin ID

D000098415;

UMLS CUI

C5940615;

Allowable qualifiers
- classification [MeSH Qualifier]
- economics [MeSH Qualifier]
- ethics [MeSH Qualifier]
- history [MeSH Qualifier]
- legislation and jurisprudence [MeSH Qualifier]
- standards [MeSH Qualifier]
- statistics and numerical data [MeSH Qualifier]
- supply and distribution [MeSH Qualifier]
- trends [MeSH Qualifier]
Currated CISMeF NLP mapping
- Convolutional Neural Networks [EFMI term]
Record concept(s)
- AlexNet [MeSH concept]
- Convolutional Neural Networks [MeSH concept]
- Densely Connected Convolutional Networks [MeSH concept]
- EfficientNet [MeSH concept]
- GoogLeNet [MeSH concept]
- LeNet-5 Convolutional Neural Network [MeSH concept]
- MobileNets [MeSH concept]
- Residual Neural Network [MeSH concept]
- VGGNet [MeSH concept]
- ZFNet [MeSH concept]
Semantic type(s)
- Intellectual Product [UMLS semantic type]

Keyword's position in hierarchy(ies) :

You can consult :

recommendations
documents concerning education
documents for patients

Artificial neural networks designed to automatically learn features through the use of sequence of layers which transform one volume to the next. Convolutional neural networks initially produce a feature map by use of convolution layers consisting of a set of learnable filters convolved with input, i.e., having smaller widths and heights and the same depth as that of input volume.
A specialized deep neural network architecture that consists of one or more convolution layers that is suited for processing grid-like data, such as images. In a convolution layer, a “filter” (window or template) slides over regions of the input image to identify low-level patterns (e.g., edges) by applying convolution (a mathematical dot operation applied to the input data). Different filters can be applied to extract different features, such as edges, textures, or curves in images. Additionally, CNNs can include pooling layers, whose function is to reduce the feature dimensionality while retaining relevant features. These convolution and pooling layers get stacked on top of each other to enable this network to build up a hierarchical understanding of patterns and makes CNNs effective at tasks like image recognition and computer vision. An important aspect of this network is its ability to conserve spatial information of the original input while still performing the feature extraction
Source: Adapted from: Gu, J., Wang, Z., Kuen, J., Ma, L., Shahroudy, A., Shuai, B., Liu, T., Wang, X., Wang, G., Cai, J., & Chen, T. (2018). Recent advances in convolutional neural networks. Pattern Recognition, 77, 354–377. https://doi.org/10.1016/j.patcog.2017.10.013External Link Disclaimer International Organization for Standardization. (2022). Information technology — Artificial intelligence — Artificial intelligence concepts and terminology (ISO/IEC 22989:2022). https://www.iso.org/standard/74296.htmlExternal Link Disclaimer Li, Z., Liu, F., Yang, W., Peng, S., & Zhou, J. (2022). A Survey of Convolutional Neural Networks: Analysis, Applications, and Prospects. IEEE Transactions on Neural Networks and Learning Systems, 33(12), 6999–7019. https://doi.org/10.1109/tnnls.2021.3084827