Digital Eprints Services at Vignan's Foundation for Science, Technology & Research
Not a member yet
721 research outputs found
Sort by
Facial Emotion Recognition using DCNN Algorithm
Get PDFFacial emotion recognition (FER) is critical for human-computer interaction in areas like clinical practice and behavioral description. With the heterogeneity of human faces and kinds of
images like various facial poses such as happy, angry, sad, fear, disgust, surprised etc. and lighting, accurate and robust FER by computer models remains a challenge. Deep learning
models, particularly Deep Convolutional Neural Networks (DCNNs), have shown great promise among all FER techniques due to their powerful automatic feature extraction and computational of efficiency. On the FER2013 dataset, the
highest single-network classification accuracy has been attained in this paper. The VGGNet architecture is used, its hyper parameters are fine-tuned, and different optimization techniques are performed. This proposed model has
achieved the state-of-the-art single-network accuracy of 90% on FER2013 without using any additional training data
Predict the Face-Mask-and-Social-Distance Identification by Using Yolo and CNN
Get PDFAbstract: COVID-19 virus is still a source of concern and hazard in today's world. With such a huge population traveling, manual monitoring of social distance standards is impracticable. Regarding and with a task force and resources that are insufficient to they should be administered There is a requirement for a lightweight, durable, and reliable device. This procedure is automated by a video observation device that operates 24 hours a day, seven days a week. This study offers a detailed and practical solution to the problem. Perform person detection, social distance violation detection, and social distancing violation detection. Using an item, detect faces and classify face masks. Convolution Neural Networks, detection, and grouping (CNN)is a binary classifier that is based on. YOLOv3, Density-fundamentally based spatial bunching of bundles with commotion (DBSCAN), and double are a portion of the instruments that can assist with this. Shot Face Detector (DSFD) and Binary MobileNetV2 Surveillance video assortments were utilized to prepare a classifier. This publication also includes a comparison of various facial types. Face mask detection and classification models. Finally, to make amends for the absence of a dataset within the network, a video dataset labeling technique is presented, couple with a labeled video dataset, which is utilized to evaluate the system. The framework's presentation is estimated as far as precision, F1 score, and gauge time, which should be least for practical use. On the labeled video data set, the device has an accuracy of 91.2 percent and an F1 rating of 90. Seventy-nine percent, with an average forecast time of seven.12 seconds for seventy eight frames of vide
Structured Ranking Method-based Feature Selection in Data Mining
Get PDFAbstract: Health diseases have been issued seriously harmful in
human life due to different dehydrated food and disturbance of
working environment in the organization. Precise prediction and
diagnosis of disease become a more serious and challenging task for primary deterrence, recognition, and treatment. Thus, based on the above challenges, we proposed the Medical Things (MT) and machine learning models to solve the healthcare problems with appropriate services in disease supervising, forecast, and diagnosis. We developed a prediction framework with machine learning approaches to get different categories of classification for predicted disease. The framework is designed by the fuzzy model with a decision tree to lessen the data complexity. We considered heart disease for experiments and experimental evaluation determined the prediction for categories of classification. The number of decision trees (M) with samples (MS), leaf node (ML), and learning rate (I) is determined as MS=20, ML=3, I=0.1, then mean test score(m) is 20
Time-Periodic Thermal Boundary Effects on Porous Media Saturated with Nanofluids: CGLE Model for Oscillatory Mode
Get PDFAbstract
The stability of nonlinear nanofluid convection is examined using the complex matrix differential operator theory. With the help of finite amplitude analysis, nonlinear convection in a porous medium is investigated that has been saturated with nanofluid and subjected to thermal modulation. The complex Ginzburg-Landau equation (CGLE) is used to determine the finite amplitude convection in order to evaluate heat and mass transfer. The small amplitude of convection is considered to determine heat and mass transfer through the porous medium. Thermal modulation of the system is predicted to change sinusoidally over time, as shown at the boundary. Three distinct modulations IPM, OPM, and LBMOhave been investigated and found that OPM and LBMO cases are used to regulate heat and mass transfer. Further, it is found that modulation frequency ( ω f varying from 2 to 70) reduces heat and mass transfer while modulation amplitude ( δ 1 varying from 0.1 to 0.5 ) enhances both
Dense region clustering using bayesian rose tree algorithm and depth first search with collaborating filtering
Get PDFIn this research, segmentation of an image into displace areas, which means every region fulfils a partition criterion. In this research, the problem of finding the maximum density region in an image has been resolved by applying the Gaussian-Filter. Further, it includes the BRT structure which includes the use of multiplicative algorithms for graph clustering. The BRT algorithm is mainly considered for gene analysis and optimization process. The outcomes found good in distinguishing and investigating complex natural designs utilizing chart bunching, collective sifting and profundity first inquiry. The inborn and extraneous qualities are additionally determined during quality ID investigation
Spectroscopic Properties Of Nd3+ Ions IN LIO2-MO-SB203 GLASSES
No full textThe primary objective of the present work is to study the spectroscopic properties of Nd3+ ions in Li2Osingle bondMO (M=Mg, Ca and Sr)single bondSb2O3 glass system. The current glass samples have been synthesized by the melt-quenching technique. The physical and thermal properties of the prepared glass samples have been calculated. Optical absorption, IR, photoluminescence (PL) and decay characteristics were recorded at ambient temperature. The Judd–Ofelt (J-O) intensity parameters Ω2, Ω4 and Ω6 for Nd2O3 mixed glasses were calculated using absorbance spectra and found to be the order Ω2 > Ω6 > Ω4. The SrO mixed glass sample J-O parameter Ω2 is found to be the lowest and has the highest emission intensity, radiative probability (A) and branching ratio (β) compared to other two glass samples MgO and CaO. The lifetime of 4F3/2 transition has been experimentally recorded at λexc= 585 nm and found that SrO mixed glass is a favorable one and getting the highest quantum efficiency (η = 64.5%). The chromaticity coordinates confirmed that the prepared glass samples were potential under 585 nm excitation wavelength. The IR studies suggest that the structural disorder may reduce the low phonon losses and increase the PL output in SrO mixed glasses. The studied results may useful for enhancing optical amplification as well as potential to generate in solid state NIR laser applications at 1.066 µm
Covid-19 Automatic Detection from CT Images through Transfer Learning
Get PDFDuring the last decade, several studies have been conducted to improve efficiency and robustness in the detection and segmentation of brain tumors based on different parameters like size, shape, location, and contrasts. This study proposes Multimodal Attention-gated Cascaded U-Net (MAC U-Net) model to address the performance issues observed in the detection and segmentation of low-grade tumors. The effectiveness of group normalization with attention gate is also explored with skip connections to segment small-scale brain tumors using several highlighted salient features. The model is evaluated on the brain tumor benchmark dataset BraTS2018 over various performance metrics such as Dice, IoU, Sensitivity, Specificity, and Accuracy. Experimental results illustrate that the proposed MAC U-net on BraTS 2018 dataset outperforms baseline U-nets with 94.47, 84.12, and 82.72 dice similarity coefficient values on HGG and 85.71, 78.85 and 74.16 on LGG subjects with Ground
Truth values of Complete Tumor, Tumor Core, and Enhancing tumor, respectively. The proposed model is also evaluated on BraTS 2019 and BraTS 2020 datasets. Moreover, MAC U-net achieves superior performance over typical conventional brain tumor segmentation methods especially in terms of low-grade gliomas
Hybrid Intelligence for Multimedia Data in Intra IoT (IIoT) Cloud by Persistent homology
Get PDFAbstract- Due to burgeoning demands from Internet/intranet
enabled 2D,3D image services in the public, social network
society and education/university premises is increasing at an
alarming rate. Intra Internet of Things (IIoT) cloud taken a
giant leap as one of the most promising fields with a plethora
of solution. There is a need of Topological image data
analysis (TIDA) in Intra internet of Things(IIoT) cloud with
the support of AI,ML,RNN technique due to the upcoming
container-based technology to reduce time frame, value,
censorious performance, load balancing level and
maximizing resource utilization in GPU for container-based
technology with success rate of steering the diverse set of
video/images. The proposed system develops a service for an
Image analytics for container-based technology in the Intra
IoT cloud by persistent homology
Detecting Covid19 and pneumonia from chest X-ray images using deep convolutional neural networks
Get PDFWith the intent growth of web-based data, document classification has become an important task that can be used in many real-time applications to handle and organize text documents. In the traditional approaches, text documents are encoded using fixed length feature vector representation. Compared to the static, fixed length representation, a text document can be better represented using variable length feature vector representation, where text documents are allowed to have variable number of features. In this paper, SVM-based classification is used as it does not suffer much from the curse of dimensionality. User-defined kernel functions such as Jaccard coefficient kernel, n-gram kernel, and string subsequence kernels are used to find the kernel value between a pair of documents. To prove the performance of the proposed method, benchmark datasets like Reuters-21578 and Reuters-8, the most widely used datasets for text classification, are used for our experimental studies. Based on our experimental studies,
we claim that SVM using N-gram kernel gives better performance on Reuters-21578 dataset and SVM using string
subsequence kernel gives better performance on Reuters-8 dataset. We also observed that minor modifications to the
user-defined kernel improve the performance of the model
Recurrent nonsymmetric deep auto encoder approach for network intrusion detection system
Get PDFYOLOv3, the third edition of the YOLO family, performs well on object detection, but using it for real-time vehicle and object detection on unmanned vehicles with limited computing capacity remains a very challenging task YOLOv3 has a high computational complexity. The main objective is to develop network architecture for vehicle and object detection based on YOLOv3. The total work will be broken down into three phases.
Firstly, to reduce the model size and computing complexity, we introduce L1 regularization to the batch normalization layer, which allows us to recognize and remove distracting channels and layers. Secondly, to reduce the missed detection in
crowded scenes and locate targets better, the Merge Soft-NMS which merges the bounding boxes with high overlap is designed based on Soft-NMS. Thirdly, considering the obvious aspect ratio of vehicle and objects, the anchor boxes which are designed
based on multi-class is redesigned for better vehicle and object matching and localization in YOLOv3. In the experiment, compared with SINGLE SHOT DETECTION (SSD) and YOLOv3 which performs well on detection accuracy and speed is effective and compact for vehicle and object detection