Dr sachin Gaur
@kecua.ac.in
Assistant Professor Computer Science & Engineering Department
Bipin Tripathi Kumaon Institute of Technology
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Sachin Gaur, Anirudh Kandwal, and Bhaskar Pandey
Institute of Advanced Engineering and Science
Diabetes detection is pivotal in disease management and complication prevention. Traditional screening methods, like blood tests, are invasive and time-consuming. Deep learning has emerged as a non-invasive and automated alternative for diabetes detection. Convolutional neural networks (CNNs) excel in image analysis tasks, making them ideal for this purpose. This paper employs a CNN-based method for diabetes prediction using retinal images, utilizing the DenseNet169 architecture for feature extraction and diabetic retinopathy (DR) prediction. The APTOS 2019 blindness detection dataset from Kaggle, containing around 13,000 retinal images, is used for training. Pre-processing and normalization precede feature extraction, followed by the prediction of the DR stage. The model aims to classify retinal images into five stages of DR (0 to 4), ranging from no DR to proliferative DR. Our model achieved over 82% accuracy, outperforming advanced algorithms. Model evaluation includes accuracy, precision, recall, and F1 score measures.
Pradeep Singh Rawat, Sachin Gaur, Varun Barthwal, Punti Gupta, Debjani Ghosh, Deepak Gupta, and Joel JP C. Rodrigues
Elsevier BV
Sachin Gaur, Milind Pandey, and Himanshu
Springer Science and Business Media LLC
Varun Barthwal, M. M. S. Rauthan, Rohan Varma, and Sachin Gaur
Springer Science and Business Media LLC
Sachin Gaur, Nitesh Tiwari, Sneha Vyas, and Milind Pandey
FOREX Publication
Recent developments in deep learning techniques have led to remarkable progress in facial recognition. As a component of biometric verification, human face recognition has become widely used in a variety of applications, including surveillance systems, home entry access, mobile face unlocking, and network security. Conventional facial recognition techniques are especially useful when dealing with low-resolution photos or difficult lighting situations. The K-nearest neighbor (KNN) classifier has been used in this paper. KNN is a non-parametric, instance-based learning algorithm that is commonly used for classification tasks. Principal Component Analysis (PCA) and local binary pattern (LBP) are used in this study to develop face identification. Both contrast stretching and grayscale were used to ensure ease of computation. The study was conducted on two separate and through multiple tests with different values for k, the highest accuracy obtained was at k=1 for both datasets. The smaller user dataset achieved 91% accuracy and CASIA-WebFace obtained 87% model accuracy.
Suraj Singh Panwar, M. M. S. Rauthan, Varun Barthwal, Sachin Gaur, and Nidhi Mehra
Springer Nature Singapore
Sachin Gaur, Navneet Tripathi, and Jyoti Pandey
EJournal Publishing
In the digital age, protecting the ownership and data veracity of digital documents is a major challenge. To address the issues concerning copyright protection and data verification of digital media, digital watermarking has emerged as a solution. In this paper, we aspire to make a modest contribution to this emerging and exciting field by presenting our proposed adaptive hybrid image watermarking approach that combines Discrete Wavelet Transform (DWT) and Singular Value Decomposition (SVD). Our method involves applying DWT to both the host image and watermark, followed by singular decomposition using SVD on the Low-Low (LL) component of both images. Now modify the singular values of the host image by the singular values of the watermark, and then inverse SVD is applied, followed by inverse DWT, to obtain the watermarked image. After that, the reverse process is applied to obtain the watermark image. Finally, we evaluate our approach’s performance by measuring the Peak Signal-to-Noise Ratio (PSNR) between the original and watermarked image as well as the Normalized Cross-Correlation (NCC) between the original and extracted watermark. Simulation results indicate that the proposed method is rich in terms of robustness, imperceptibility and capacity than the previously presented schemes.
Sachin Gaur, Krista Chaudhary, Vikas Goel, and Varun Barthwal
Springer Science and Business Media LLC
Pradeep Singh Rawat, Anuj Kumar Yadav, Varun Barthwal, and Sachin Gaur
IEEE
In the present era of computing, communication and Technology, natural disaster can be controlled in an efficient manner. The natural disaster in hill sate is common. It can be mitigated using machine learning techniques. In this manuscript out objective is proposed a machine learning model which focus on rainfall induced landslide prediction in uttarakhand state districts using benchmark dataset. There is a good correlation between landslide and antecedent rain fall. The antecedent rain fall supports the machine learning model for better accuracy and correctness. The machine learning model with optimal performance metrics provides the prior information about the level rainfall and its impact level on landslide in a study area of the focused state. The results show that Random Forest model outperforms the linear model, SVR model, and neural network model respectively. The key performance indicators i.e. mean absolute error(MAE), and root mean square error(RMSE) are improved by a factor of 79.05%, and 83.34% respectively. The key performance indicators evaluated and analyzed against state of art methodologies i.e. Random Forest model outperforms the linear model, SVR model, and neural network
Amit Kumar and Sachin Gaur
IEEE
As the measurements received from RTUs to the Control Center are transmitted via a transmission medium e.g. telephone, fibre optics, wireless medium, it is not possible that the data transmitted is 100% error free always. Other reasons for receiving bad measurements at control centers may be due to wrong reading of the meter. There can be a number of reasons for measurement’s value to be recorded as wrong, e.g. outage of meter, drift in meter and bias in the meter.Therefore the measurements received may be erroneous sometimes, due to which the state estimation results may be misleading, and consequently can cause problem in monitoring and control of power system. Thus it is necessary to remove bad data from the measurement set or establish some robust state estimation techniques which can remove the effects of bad data on the estimated states.In this paper the problem of multiple bad measurements detection and identification is defined as a binary variables optimization problem and it’s solutions are obtained by using Binary Particle Swarm Optimization (BPSO). It is observed that this method can be used to identify multiple interacting erroneous measurements.
Digital image watermarking has been proposed to protect the digital multimedia content. The main objectives of watermarking scheme are robustness, reliability, security against numerous attacks. To improve the imperceptibility, robustness and capacity of the watermarked image, this paper presents a transform domain watermarking method using spatial frequency and block SVD. The spatial frequency is used to select the appropriate blocks for embedding the watermark image by transforming the SVD coefficients of these blocks of the cover image. In this paper first we scramble the cover image by ZIG-ZAG sequencing and then rearranged. After that Shift Invariant Discrete Wavelet Transformed (SIDWT) cover image is partioned in to non-overlapping blocks. Then find out the spatial frequency of these blocks, those blocks which spatial frequency value greater than threshold value are selected for embedding process. Now the watermark image directly embedded by modifying the SVD coefficient of these blocks and get watermarked image. Then inverse process is applied for extracting for watermark image form noisy image. Experimental outcomes show that the proposed scheme is higher imperceptible, robust against various image processing attacks and produce improved results as compared to previous presented schemes
Sachin Gaur and Vinay Kumar Srivastava
IEEE
In the present scenario Digital image watermarking is a powerful method for solving the problems of tamper detection, rightful ownership, copyright protection and content authentication. In this papera secure hybrid digital image watermarking scheme based on Redundant Discrete wavelet transform (RDWT), Discrete Cosine Transform (DCT) and Singular value decomposition (SVD) in zigzag order with Arnold transform is presented. Watermark image is scrambled by Arnold transform to boost up its secrecy and robustness. In presented scheme, a gray scale cover image is rearranged through zigzag sequence and then RDWT is implemented on this reordered cover image. After that DCT, SVD is implemented on mid and high frequency sub-bands (LH, LH, and HH) of cover image and modified the singular values of these sub-bands by embedding the scrambled gray scale watermark image. This presented scheme is more imperceptible and an enormous capacity due to the properties of RDWT and SVD. The benefit of the presented schemeis more robust and secured against various image processing attacks. Analysis and experimental outcomes show that the presented scheme is rich in terms of imperceptibility, robustness, capacity and security from earlier proposed schemes.
Sachin Gaur and Vinay Kumar Srivastava
IEEE
Digital image Watermarking gives an efficient method for copyright protection. In this paper a robust and secure algorithm of watermarking based on Redundant Discrete wavelet transform (RDWT), Singular value decomposition (SVD) and Improve Arnold transform is presented. The watermark image is scrambled by Improved Arnold transform to boost up its confidentiality and robustness. In the proposed scheme, after applying RDWT and SVD to each sub-band of the gray scale host image, we modify the singular values of the host image by embedding the gray scale scrambled watermark image. This presented method is more imperceptible and has an extensive capacity due to SVD and RDWT. The advantage of the given method is that it is highly robust against various image processing attacks. Analysis and experimental results demonstrate that the proposed scheme performs better in comparison previously introduced RDWT-SVD based method.