Novel CNN with investigation on accuracy by modifying stride, padding, kernel size and filter numbers

Naseri, HamidReza; Mehrdad, Vahid

doi:10.1007/s11042-023-14603-x

Cited by 13 publications

(4 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, many other hyperparameters, including kernel sizes, strides, padding, activation functions, and kernel initializer, are utilized by Huang and Jafari 28 that proposed CDRAGAN and BAGAN-GP, which are the state-of-the-art class conditional GAN methods. Furthermore, the number of kernels is determined as two to the powers, including 32, 64, and 128, commonly used in the existing studies using convolutional neural networks 48 , 49 .…”

Section: Methodsmentioning

confidence: 99%

Imbalanced spectral data analysis using data augmentation based on the generative adversarial network

Chung,

Zhang,

Saimon

et al. 2024

Sci Rep

View full text Add to dashboard Cite

Spectroscopic techniques generate one-dimensional spectra with distinct peaks and specific widths in the frequency domain. These features act as unique identities for material characteristics. Deep neural networks (DNNs) has recently been considered a powerful tool for automatically categorizing experimental spectra data by supervised classification to evaluate material characteristics. However, most existing work assumes balanced spectral data among various classes in the training data, contrary to actual experiments, where the spectral data is usually imbalanced. The imbalanced training data deteriorates the supervised classification performance, hindering understanding of the phase behavior, specifically, sol-gel transition (gelation) of soft materials and glycomaterials. To address this issue, this paper applies a novel data augmentation method based on a generative adversarial network (GAN) proposed by the authors in their prior work. To demonstrate the effectiveness of the proposed method, the actual imbalanced spectral data from Pluronic F-127 hydrogel and Alpha-Cyclodextrin hydrogel are used to classify the phases of data. Specifically, our approach improves 8.8%, 6.4%, and 6.2% of the performance of the existing data augmentation methods regarding the classifier’s F-score, Precision, and Recall on average, respectively. Specifically, our method consists of three DNNs: the generator, discriminator, and classifier. The method generates samples that are not only authentic but emphasize the differentiation between material characteristics to provide balanced training data, improving the classification results. Based on these validated results, we expect the method’s broader applications in addressing imbalanced measurement data across diverse domains in materials science and chemical engineering.

show abstract

Section: Methodsmentioning

confidence: 99%

Imbalanced spectral data analysis using data augmentation based on the generative adversarial network

Chung,

Zhang,

Saimon

et al. 2024

Sci Rep

View full text Add to dashboard Cite

show abstract

“…Padding, stride, and filters are important parameters in CNNs that determine the size and resolution of the output feature maps (Feat_Maps) [86][87][88][89][90][91][92][93]. These parameters play a crucial role in controlling the amount of information that is retained in the Feat_Maps and can greatly impact the performance of the network.…”

Section: Padding Stride and Filtersmentioning

confidence: 99%

Convolutional Neural Networks: A Survey

Krichen

2023

Computers

154

View full text Add to dashboard Cite

Artificial intelligence (AI) has become a cornerstone of modern technology, revolutionizing industries from healthcare to finance. Convolutional neural networks (CNNs) are a subset of AI that have emerged as a powerful tool for various tasks including image recognition, speech recognition, natural language processing (NLP), and even in the field of genomics, where they have been utilized to classify DNA sequences. This paper provides a comprehensive overview of CNNs and their applications in image recognition tasks. It first introduces the fundamentals of CNNs, including the layers of CNNs, convolution operation (Conv_Op), Feat_Maps, activation functions (Activ_Func), and training methods. It then discusses several popular CNN architectures such as LeNet, AlexNet, VGG, ResNet, and InceptionNet, and compares their performance. It also examines when to use CNNs, their advantages and limitations, and provides recommendations for developers and data scientists, including preprocessing the data, choosing appropriate hyperparameters (Hyper_Param), and evaluating model performance. It further explores the existing platforms and libraries for CNNs such as TensorFlow, Keras, PyTorch, Caffe, and MXNet, and compares their features and functionalities. Moreover, it estimates the cost of using CNNs and discusses potential cost-saving strategies. Finally, it reviews recent developments in CNNs, including attention mechanisms, capsule networks, transfer learning, adversarial training, quantization and compression, and enhancing the reliability and efficiency of CNNs through formal methods. The paper is concluded by summarizing the key takeaways and discussing the future directions of CNN research and development.

show abstract

“…In this study, we developed an enhanced U-Net model with several improvements [13]. First, we streamlined the model by removing redundant crops, which not only reduced computational complexity but also improved efficiency [14,15]. Second, we simplified the model by decreasing the number of epochs, resulting in faster training while maintaining performance [16].…”

Section: Modelmentioning

confidence: 99%

A simplified U-net model for Covid-19 CT image segmentation

Zhan

2023

ACE

View full text Add to dashboard Cite

The ongoing COVID-19 pandemic has highlighted the importance of accurate and efficient medical image analysis to aid in the diagnosis and treatment of patients. In particular, the segmentation of COVID-19 medical images has become a critical task to identify regions of interest, such as the infected lung areas, and to track disease progression. Traditional image segmentation methods have been widely used in medical image analysis. However, these methods are often challenged by the complex and diverse nature of COVID-19 images, as well as the limited availability of data. In this paper, we propose a simplified version of the U-Net that eliminates redundant crop operations. This simplification reduces computational complexity and memory usage, and enables the model to learn from larger input images, resulting in better performance. We evaluate the performance of our simplified U-Net model on a public COVID-19 dataset and demonstrate that our model achieves state-of-the-art results while using fewer computational resources.

show abstract

Novel CNN with investigation on accuracy by modifying stride, padding, kernel size and filter numbers

Cited by 13 publications

References 27 publications

Imbalanced spectral data analysis using data augmentation based on the generative adversarial network

Imbalanced spectral data analysis using data augmentation based on the generative adversarial network

Convolutional Neural Networks: A Survey

A simplified U-net model for Covid-19 CT image segmentation

Contact Info

Product

Resources

About