Empirical Evaluation of Activation Functions in Deep Convolution Neural Network for Facial Expression Recognition

Khalid, Muhammad Irfan; Baber, Junaid; Kasi, Mumraiz Khan; Bakhtyar, Maheen; Devi, V. Ajantha; Sheikh, Naveed

doi:10.1109/tsp49548.2020.9163446

Cited by 33 publications

(18 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A sigmoid function is applied to the output of the last layer to get a probability value. The activation in the MLP was Leaky ReLU (19). A Dropout rate of 0.1 was set between layers.…”

Section: Mlp Classifiersmentioning

confidence: 99%

Contribution of T Cell Receptor Alpha and Beta CDR3, MHC Typing, V and J Genes to Peptide Binding Prediction

2021

View full text Add to dashboard Cite

IntroductionPredicting the binding specificity of T Cell Receptors (TCR) to MHC-peptide complexes (pMHCs) is essential for the development of repertoire-based biomarkers. This affinity may be affected by different components of the TCR, the peptide, and the MHC allele. Historically, the main element used in TCR-peptide binding prediction was the Complementarity Determining Region 3 (CDR3) of the beta chain. However, recently the contribution of other components, such as the alpha chain and the other V gene CDRs has been suggested. We use a highly accurate novel deep learning-based TCR-peptide binding predictor to assess the contribution of each component to the binding.MethodsWe have previously developed ERGO-I (pEptide tcR matchinG predictiOn), a sequence-based T-cell receptor (TCR)-peptide binding predictor that employs natural language processing (NLP) -based methods. We improved it to create ERGO-II by adding the CDR3 alpha segment, the MHC typing, V and J genes, and T cell type (CD4+ or CD8+) as to the predictor. We then estimate the contribution of each component to the prediction.Results and DiscussionERGO-II provides for the first time high accuracy prediction of TCR-peptide for previously unseen peptides. For most tested peptides and all measures of binding prediction accuracy, the main contribution was from the beta chain CDR3 sequence, followed by the beta chain V and J and the alpha chain, in that order. The MHC allele was the least contributing component. ERGO-II is accessible as a webserver at http://tcr2.cs.biu.ac.il/ and as a standalone code at https://github.com/IdoSpringer/ERGO-II.

show abstract

“…A sigmoid function is applied to the output of the last layer to get a probability value. The activation in the MLP was Leaky ReLU (19). A Dropout rate of 0.1 was set between layers.…”

Section: Mlp Classifiersmentioning

confidence: 99%

Contribution of T Cell Receptor Alpha and Beta CDR3, MHC Typing, V and J Genes to Peptide Binding Prediction

2021

View full text Add to dashboard Cite

show abstract

“…The structure of the neural network used in our experiment is a structure with four hidden layers, each hidden layer is composed of n neurons, where n is grid searched from the value set f64; 128; 256g, each hidden layer is activated with ReLU function. 42 In the output layer, the SoftMax activation function is performed to get the probabilistic prediction. Learning rate, which is an important parameter to the final predictive performance, is determined by grid searching from the space of f10 −4 ; 10 −3 ; 10 −2 g.…”

Section: Resultsmentioning

confidence: 99%

Multi-spectral cloud detection based on a multi-dimensional and multi-grained dense cascade forest

Shao

Zou

2021

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

Cloud detection in satellite images is a vital step for cloud/land recognition, cloud/ snow discrimination, and cloud shadow removal. Accurate cloud detection plays an important role in land resource management, environmental pollution monitoring, and land target recognition. Deep learning (DL) algorithms have shown great progress in cloud detection. However, as the complexity of the DL-based model increases, cloud detection efficiency decreases. DLbased cloud detection models are unable to successfully balance the performance-efficiency tradeoff. In our study, a multi-dimensional and multi-grained dense cascade forest (MDForest) is proposed for multi-spectral cloud detection. MDForest is a deep forest structure that automatically extracts low-level and high-level features from satellite cloud images end-to-end; a multi-dimensional and multi-grained scanning mechanism is introduced to capture the spectral information of multi-spectral satellite images while enhancing the representation learning ability of cascade forest. The experimental results on the HJ-1A/1B dataset show that MDForest improves the performance of cloud detection and possesses a good inference efficiency compared with DL-based cloud detection methods, which makes the proposed MDForest satisfy the application where good performance and high efficiency are both required. © The Authors.Published by SPIE under a Creative Commons Attribution 4.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

show abstract

“…The model consists of six bidimensional convolutional layers in which the stride is fixed at 2 in each dimension to downsample the feature maps size after each convolution. The activation function of the first five layers is the LeakyReLu [ 52 ] to avoid blockage during training due to negative values in the input signals. In layers 2 to 5, batch normalization is also applied over every instance.…”

Section: Methodsmentioning

confidence: 99%

Video-Rate Quantitative Phase Imaging Using a Digital Holographic Microscope and a Generative Adversarial Network

Guerrero-Castañeda

Trujillo

Doblas

2021

Sensors

View full text Add to dashboard Cite

The conventional reconstruction method of off-axis digital holographic microscopy (DHM) relies on computational processing that involves spatial filtering of the sample spectrum and tilt compensation between the interfering waves to accurately reconstruct the phase of a biological sample. Additional computational procedures such as numerical focusing may be needed to reconstruct free-of-distortion quantitative phase images based on the optical configuration of the DHM system. Regardless of the implementation, any DHM computational processing leads to long processing times, hampering the use of DHM for video-rate renderings of dynamic biological processes. In this study, we report on a conditional generative adversarial network (cGAN) for robust and fast quantitative phase imaging in DHM. The reconstructed phase images provided by the GAN model present stable background levels, enhancing the visualization of the specimens for different experimental conditions in which the conventional approach often fails. The proposed learning-based method was trained and validated using human red blood cells recorded on an off-axis Mach–Zehnder DHM system. After proper training, the proposed GAN yields a computationally efficient method, reconstructing DHM images seven times faster than conventional computational approaches.

show abstract

Empirical Evaluation of Activation Functions in Deep Convolution Neural Network for Facial Expression Recognition

Cited by 33 publications

References 12 publications

Contribution of T Cell Receptor Alpha and Beta CDR3, MHC Typing, V and J Genes to Peptide Binding Prediction

Contribution of T Cell Receptor Alpha and Beta CDR3, MHC Typing, V and J Genes to Peptide Binding Prediction

Multi-spectral cloud detection based on a multi-dimensional and multi-grained dense cascade forest

Video-Rate Quantitative Phase Imaging Using a Digital Holographic Microscope and a Generative Adversarial Network

Contact Info

Product

Resources

About