Robust Drowsiness Detection for Vehicle Driver using Deep Convolutional Neural Network

Saif, A F M Saifuddin; Mahayuddin, Zainal Rasyid

doi:10.14569/ijacsa.2020.0111043

Cited by 17 publications

(9 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared with semantic segmentation methods such as PspNet, DeconNet, DeepLabv3+ and GCRF, the results using attention mechanism and feature fusion are 4.08%, 11.86%, 3.14% and 13.41% higher, respectively. It is 1.45% higher than the current SOTA method [35].…”

Section: A Experimental Results Analysis Of Image Semantic Segmentati...mentioning

confidence: 59%

Research on Image Semantic Segmentation Based on Hybrid Cascade Feature Fusion and Detailed Attention Mechanism

Du,

Liang

2024

IEEE Access

View full text Add to dashboard Cite

In view of the low segmentation accuracy for small-scale object and insufficient segmentation of local boundary for semantic segmentation methods based on Deep Learning, this paper proposes an image semantic segmentation approach based on attention mechanism and feature fusion. On the basis of ensuring the overall accuracy, the segmentation accuracy of small-scale object and local boundary is improved, and it meets the requirement of accurately segmenting the object in the complex background. Firstly, an image semantic segmentation model based on hybrid cascade and feature fusion is proposed, and the hybrid concatenation and multi-cores pooling methods are used to extract deeper semantic information. Then, a cross-stages fusion approach is designed to divide the backbone network of the encoder stage in the network and the improved Atrous Spatial Pyramid Pooling module into three stages to fully utilize the different semantic information of the shallow and deep layers. Thirdly, the attention mechanism is introduced into the hybrid cascade and feature fusion image semantic segmentation network model, and image semantic segmentation model based on cross-stages and attention mechanisms is explained. Self attention is added to channel attention enhances the connection between feature maps, and one-dimensional convolution in the spatial attention mechanism is used to increase the spatial receptive field. The final results on the public dataset PASCAL VOC2012 and SUIM show that MIoUs have reached 86.68% and 61.55% respectively, and it proved that the overall accuracy of the approach proposed in this paper is higher than other ones.INDEX TERMS Semantic segmentation, hybrid cascade feature fusion, detailed attention mechanism, deep learning.

show abstract

Section: A Experimental Results Analysis Of Image Semantic Segmentati...mentioning

confidence: 59%

Research on Image Semantic Segmentation Based on Hybrid Cascade Feature Fusion and Detailed Attention Mechanism

Du,

Liang

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Usage of multiple datasets is common for researchers using different parameters and including or excluding different levels of details while annotating the data [34,[51][52][53][59][60][61][62][63]. or FreiHAND, researchers performed cross-dataset generalization to achieve improved results [35].…”

Section: Step Wise Framework Analysismentioning

confidence: 99%

Vision based 3D Gesture Tracking using Augmented Reality and Virtual Reality for Improved Learning Applications

Mahayuddin¹,

Saif²

2021

IJACSA

Self Cite

View full text Add to dashboard Cite

3D gesture recognition and tracking based augmented reality and virtual reality have become a big interest of research because of advanced technology in smartphones. By interacting with 3D objects in augmented reality and virtual reality, users get better understanding of the subject matter where there have been requirements of customized hardware support and overall experimental performance needs to be satisfactory. This research investigates currently various vision based 3D gestural architectures for augmented reality and virtual reality. The core goal of this research is to present analysis on methods, frameworks followed by experimental performance on recognition and tracking of hand gestures and interaction with virtual objects in smartphones. This research categorized experimental evaluation for existing methods in three categories, i.e. hardware requirement, documentation before actual experiment and datasets. These categories are expected to ensure robust validation for practical usage of 3D gesture tracking based on augmented reality and virtual reality. Hardware set up includes types of gloves, fingerprint and types of sensors. Documentation includes classroom setup manuals, questionaries, recordings for improvement and stress test application. Last part of experimental section includes usage of various datasets by existing research. The overall comprehensive illustration of various methods, frameworks and experimental aspects can significantly contribute to 3D gesture recognition and tracking based augmented reality and virtual reality.

show abstract

“…With the great breakthrough of the deep learning theory, it has been popularly applied in the medical imaging processing domain, such as the image U-net model for the superresolution [27], the convolutional long short-term deep network for recognition of human action [28], and the graph resnet for motor imagery classification [29], as well as the image denoising. For example, K. Zhang [33].…”

Section: Introductionmentioning

confidence: 99%

The Medical Image Denoising Method Based on the CycleGAN and the Complex Shearlet Transform

Liu,

Huang,

Tahir

et al. 2023

IJACSA

View full text Add to dashboard Cite

Medical image denoising plays an important role for the noise in the medical images can reduce the visibility, thereby affecting the diagnostic results of the doctors. Although good results have been achieved by the well-known deep learning-based denoising methods for their strong ability of learning, the loss of structural feature information and the well preservation of the edge information have not attracted considerable attention. To deal with these problems, a novel medical image denoising method based on the improved CycleGAN and the complex shearlet transform(CST) is proposed. The CST is used to construct the generator to embed more feature information in the training process and the denoising process is modeled to adversarial learn the mapping between the noise-free image domain and the noisy image domain. With the mechanism of the recurrent learning from the CycleGAN, the proposed method does not need the paired training data, which obviously speeds up the training and is more convenient than other classical methods. By comparing with five state-of-the-art denoising methods, experiments on the open dataset fully prove the accuracy and efficiency of the proposed method in terms of the visual quality and the quantitative PSNR, SSIM, and EPI.

show abstract

Robust Drowsiness Detection for Vehicle Driver using Deep Convolutional Neural Network

Cited by 17 publications

References 31 publications

Research on Image Semantic Segmentation Based on Hybrid Cascade Feature Fusion and Detailed Attention Mechanism

Research on Image Semantic Segmentation Based on Hybrid Cascade Feature Fusion and Detailed Attention Mechanism

Vision based 3D Gesture Tracking using Augmented Reality and Virtual Reality for Improved Learning Applications

The Medical Image Denoising Method Based on the CycleGAN and the Complex Shearlet Transform

Contact Info

Product

Resources

About