-
Efficient Human Pose Estimation: Leveraging Advanced Techniques with MediaPipe
Authors:
Sandeep Singh Sengar,
Abhishek Kumar,
Owen Singh
Abstract:
This study presents significant enhancements in human pose estimation using the MediaPipe framework. The research focuses on improving accuracy, computational efficiency, and real-time processing capabilities by comprehensively optimising the underlying algorithms. Novel modifications are introduced that substantially enhance pose estimation accuracy across challenging scenarios, such as dynamic m…
▽ More
This study presents significant enhancements in human pose estimation using the MediaPipe framework. The research focuses on improving accuracy, computational efficiency, and real-time processing capabilities by comprehensively optimising the underlying algorithms. Novel modifications are introduced that substantially enhance pose estimation accuracy across challenging scenarios, such as dynamic movements and partial occlusions. The improved framework is benchmarked against traditional models, demonstrating considerable precision and computational speed gains. The advancements have wide-ranging applications in augmented reality, sports analytics, and healthcare, enabling more immersive experiences, refined performance analysis, and advanced patient monitoring. The study also explores the integration of these enhancements within mobile and embedded systems, addressing the need for computational efficiency and broader accessibility. The implications of this research set a new benchmark for real-time human pose estimation technologies and pave the way for future innovations in the field. The implementation code for the paper is available at https://github.com/avhixd/Human_pose_estimation.
△ Less
Submitted 13 July, 2024; v1 submitted 21 June, 2024;
originally announced June 2024.
-
VigilEye -- Artificial Intelligence-based Real-time Driver Drowsiness Detection
Authors:
Sandeep Singh Sengar,
Aswin Kumar,
Owen Singh
Abstract:
This study presents a novel driver drowsiness detection system that combines deep learning techniques with the OpenCV framework. The system utilises facial landmarks extracted from the driver's face as input to Convolutional Neural Networks trained to recognise drowsiness patterns. The integration of OpenCV enables real-time video processing, making the system suitable for practical implementation…
▽ More
This study presents a novel driver drowsiness detection system that combines deep learning techniques with the OpenCV framework. The system utilises facial landmarks extracted from the driver's face as input to Convolutional Neural Networks trained to recognise drowsiness patterns. The integration of OpenCV enables real-time video processing, making the system suitable for practical implementation. Extensive experiments on a diverse dataset demonstrate high accuracy, sensitivity, and specificity in detecting drowsiness. The proposed system has the potential to enhance road safety by providing timely alerts to prevent accidents caused by driver fatigue. This research contributes to advancing real-time driver monitoring systems and has implications for automotive safety and intelligent transportation systems. The successful application of deep learning techniques in this context opens up new avenues for future research in driver monitoring and vehicle safety. The implementation code for the paper is available at https://github.com/LUFFY7001/Driver-s-Drowsiness-Detection.
△ Less
Submitted 21 June, 2024;
originally announced June 2024.
-
BetterNet: An Efficient CNN Architecture with Residual Learning and Attention for Precision Polyp Segmentation
Authors:
Owen Singh,
Sandeep Singh Sengar
Abstract:
Colorectal cancer contributes significantly to cancer-related mortality. Timely identification and elimination of polyps through colonoscopy screening is crucial in order to decrease mortality rates. Accurately detecting polyps in colonoscopy images is difficult because of the differences in characteristics such as size, shape, texture, and similarity to surrounding tissues. Current deep-learning…
▽ More
Colorectal cancer contributes significantly to cancer-related mortality. Timely identification and elimination of polyps through colonoscopy screening is crucial in order to decrease mortality rates. Accurately detecting polyps in colonoscopy images is difficult because of the differences in characteristics such as size, shape, texture, and similarity to surrounding tissues. Current deep-learning methods often face difficulties in capturing long-range connections necessary for segmentation. This research presents BetterNet, a convolutional neural network (CNN) architecture that combines residual learning and attention methods to enhance the accuracy of polyp segmentation. The primary characteristics encompass (1) a residual decoder architecture that facilitates efficient gradient propagation and integration of multiscale features. (2) channel and spatial attention blocks within the decoder block to concentrate the learning process on the relevant areas of polyp regions. (3) Achieving state-of-the-art performance on polyp segmentation benchmarks while still ensuring computational efficiency. (4) Thorough ablation tests have been conducted to confirm the influence of architectural components. (5) The model code has been made available as open-source for further contribution. Extensive evaluations conducted on datasets such as Kvasir-SEG, CVC ClinicDB, Endoscene, EndoTect, and Kvasir-Sessile demonstrate that BetterNets outperforms current SOTA models in terms of segmentation accuracy by significant margins. The lightweight design enables real-time inference for various applications. BetterNet shows promise in integrating computer-assisted diagnosis techniques to enhance the detection of polyps and the early recognition of cancer. Link to the code: https://github.com/itsOwen/BetterNet
△ Less
Submitted 5 May, 2024;
originally announced May 2024.
-
Effect of event classifiers on jet quenching-like signatures in high-multiplicity $p+p$ collisions at $\sqrt{s} = 13$ TeV
Authors:
Hushnud Hushnud,
Omveer Singh,
Srikanta Kumar Tripathy,
Aditya Nath Mishra,
Kalyan Dey
Abstract:
The motivation behind exploring jet quenching-like phenomena in small systems arises from the experimental observation of heavy-ion-like behavior of particle production in high-multiplicity proton-proton ($p+p$) collisions. Quantifying the jet quenching in $p+p$ collisions is a challenging task, as the magnitude of the nuclear modification factor ($R_{\rm AA}$ or $R_{\rm CP}$), which is used to qu…
▽ More
The motivation behind exploring jet quenching-like phenomena in small systems arises from the experimental observation of heavy-ion-like behavior of particle production in high-multiplicity proton-proton ($p+p$) collisions. Quantifying the jet quenching in $p+p$ collisions is a challenging task, as the magnitude of the nuclear modification factor ($R_{\rm AA}$ or $R_{\rm CP}$), which is used to quantify jet quenching, is influenced by several factors, such as the estimation of centrality and the scaling factor. The most common method of centrality estimation employed by the ALICE collaboration is based on measuring charged-particle multiplicity with the V0 detector situated at the forward rapidity. This technique of centrality estimation makes the event sample biased towards hard processes like multijet final states. This bias of the V0 detector towards hard processes makes it difficult to study the jet quenching effect in high-multiplicity $p+p$ collisions. In the present article, we propose to explore the use of a new and robust event classifier, flattenicity which is sensitive to both the multiple soft partonic interactions and hard processes. The $\mathcal{P}_{\rm CP}$, a quantity analogous to $R_{\rm CP}$, has been estimated for high-multiplicity $p+p$ collisions at $\sqrt{s} = 13$ TeV using \texttt{PYTHIA8} model for both the V0M (the multiplicity classes selected based on V0 detector acceptance) as well as flattenicity. The evolution of $\mathcal{P}_{\rm CP}$ with $p_{\rm T}$ shows a heavy-ion-like effect for flattencity which is attributed to the selection of softer transverse momentum particles in high-multiplicity $p+p$ collisions.
△ Less
Submitted 12 November, 2023; v1 submitted 18 September, 2023;
originally announced September 2023.
-
Sparse Bayesian Lasso via a Variable-Coefficient $\ell_1$ Penalty
Authors:
Nathan Wycoff,
Ali Arab,
Katharine M. Donato,
Lisa O. Singh
Abstract:
Modern statistical learning algorithms are capable of amazing flexibility, but struggle with interpretability. One possible solution is sparsity: making inference such that many of the parameters are estimated as being identically 0, which may be imposed through the use of nonsmooth penalties such as the $\ell_1$ penalty. However, the $\ell_1$ penalty introduces significant bias when high sparsity…
▽ More
Modern statistical learning algorithms are capable of amazing flexibility, but struggle with interpretability. One possible solution is sparsity: making inference such that many of the parameters are estimated as being identically 0, which may be imposed through the use of nonsmooth penalties such as the $\ell_1$ penalty. However, the $\ell_1$ penalty introduces significant bias when high sparsity is desired. In this article, we retain the $\ell_1$ penalty, but define learnable penalty weights $λ_p$ endowed with hyperpriors. We start the article by investigating the optimization problem this poses, developing a proximal operator associated with the $\ell_1$ norm. We then study the theoretical properties of this variable-coefficient $\ell_1$ penalty in the context of penalized likelihood. Next, we investigate application of this penalty to Variational Bayes, developing a model we call the Sparse Bayesian Lasso which allows for behavior qualitatively like Lasso regression to be applied to arbitrary variational models. In simulation studies, this gives us the Uncertainty Quantification and low bias properties of simulation-based approaches with an order of magnitude less computation. Finally, we apply our methodology to a Bayesian lagged spatiotemporal regression model of internal displacement that occurred during the Iraqi Civil War of 2013-2017.
△ Less
Submitted 12 May, 2023; v1 submitted 9 November, 2022;
originally announced November 2022.
-
k^{th} order Slant Hankel Operators on the Polydisk
Authors:
M. P. Singh,
Oinam Nilbir Singh
Abstract:
In this paper, we initiate the notion of k^{th} order slant Hankel operators on L^2(T^n) for k greater than or equal to 2 and n greater than or equal to 1 where T^n denotes the n-torus. We give the necessary and sufficient condition for a bounded operator on L^2(T^n) to be a k^{th} order slant Hankel and discuss their commutative, compactness, hyponormal and isometric property.
In this paper, we initiate the notion of k^{th} order slant Hankel operators on L^2(T^n) for k greater than or equal to 2 and n greater than or equal to 1 where T^n denotes the n-torus. We give the necessary and sufficient condition for a bounded operator on L^2(T^n) to be a k^{th} order slant Hankel and discuss their commutative, compactness, hyponormal and isometric property.
△ Less
Submitted 18 February, 2022;
originally announced February 2022.
-
Pseudo-isotopies and diffeomorphisms of 4-manifolds
Authors:
Oliver Singh
Abstract:
A diffeomorphism $f$ of a compact manifold $X$ is pseudo-isotopic to the identity if there is a diffeomorphism $F$ of $X\times I$ which restricts to $f$ on $X\times 1$, and which restricts to the identity on $X\times 0$ and $\partial X\times I$. We construct examples of diffeomorphisms of 4-manifolds which are pseudo-isotopic but not isotopic to the identity. To do so, we further understanding of…
▽ More
A diffeomorphism $f$ of a compact manifold $X$ is pseudo-isotopic to the identity if there is a diffeomorphism $F$ of $X\times I$ which restricts to $f$ on $X\times 1$, and which restricts to the identity on $X\times 0$ and $\partial X\times I$. We construct examples of diffeomorphisms of 4-manifolds which are pseudo-isotopic but not isotopic to the identity. To do so, we further understanding of which elements of the "second pseudo-isotopy obstruction", defined by Hatcher and Wagoner, can be realised by pseudo-isotopies of 4-manifolds. We also prove that all elements of the first and second pseudo-isotopy obstructions can be realised after connected sums with copies of $S^2\times S^2$.
△ Less
Submitted 14 November, 2022; v1 submitted 30 November, 2021;
originally announced November 2021.
-
Commissioning and testing of pre-series triple GEM prototypes for CBM-MuCh in the mCBM experiment at the SIS18 facility of GSI
Authors:
A. Kumar,
A. Agarwal,
S. Chatterjee,
S. Chattopadhyay,
A. K. Dubey,
C. Ghosh,
E. Nandy,
V. Negi,
S. K. Prasad,
J. Saini,
V. Singhal,
O. Singh,
G. Sikder,
J. de Cuveland,
I. Deppner,
D. Emschermann,
V. Friese,
J. Frühauf,
M. Gumiński,
N. Herrmann,
D. Hutter,
M. Kis,
J. Lehnert,
P. -A. Loizeau,
C. J. Schmidt
, et al. (3 additional authors not shown)
Abstract:
Large area triple GEM chambers will be employed in the first two stations of the MuCh system of the CBM experiment at the upcoming Facility for Antiproton and Ion Research FAIR in Darmstadt/Germany. The GEM detectors have been designed to take data at an unprecedented interaction rate (up to 10 MHz) in nucleus-nucleus collisions in CBM at FAIR. Real-size trapezoidal modules have been installed in…
▽ More
Large area triple GEM chambers will be employed in the first two stations of the MuCh system of the CBM experiment at the upcoming Facility for Antiproton and Ion Research FAIR in Darmstadt/Germany. The GEM detectors have been designed to take data at an unprecedented interaction rate (up to 10 MHz) in nucleus-nucleus collisions in CBM at FAIR. Real-size trapezoidal modules have been installed in the mCBM experiment and tested in nucleus-nucleus collisions at the SIS18 beamline of GSI as a part of the FAIR Phase-0 program. In this report, we discuss the design, installation, commissioning, and response of these GEM modules in detail. The response has been studied using the free-streaming readout electronics designed for the CBM-MuCh and CBM-STS detector system. In free-streaming data, the first attempt on an event building based on the timestamps of hits has been carried out, resulting in the observation of clear spatial correlations between the GEM modules in the mCBM setup for the first time. Accordingly, a time resolution of $\sim$15\,ns have been obtained for the GEM detectors.
△ Less
Submitted 12 August, 2021;
originally announced August 2021.
-
A Study of Multifractal Analysis in 16O-AgBr Collisions at 60A and 200A GeV
Authors:
Nazeer Ahmad,
Tufail Ahmad,
Omveer Singh,
Shakeel Ahmad
Abstract:
A multifractal analysis to study the multiparticle dynamics in 60A and 200A GeV/c 16O-AgBr collisions has been performed in the pseudorapidity phase space. Multifractal moments Gq as the function of pseudorapidity bin size for different order of the moments, q have been calculated. The power-law behaviour has been observed in the considered data sets. The variation of multifractal dimensions, Dq a…
▽ More
A multifractal analysis to study the multiparticle dynamics in 60A and 200A GeV/c 16O-AgBr collisions has been performed in the pseudorapidity phase space. Multifractal moments Gq as the function of pseudorapidity bin size for different order of the moments, q have been calculated. The power-law behaviour has been observed in the considered data sets. The variation of multifractal dimensions, Dq and multifractal spectral function, f($α$q) with order of the moments, q have been studied thoroughly. Dq is found to decrease with increasing order of the moments, q indicating thereby a self-similar behaviour in the multiparticle production in the considered collisions. We have also found a concave downward curve of multifractal spectral function with maxima q=0.
△ Less
Submitted 10 July, 2020;
originally announced August 2020.
-
Dynamics of mosquito swarms over a moving marker
Authors:
Puneet Jain,
Om Prakash Singh,
Sachit Butail
Abstract:
Insect swarms are a model system for understanding collective behavior where the collective motion appears in disorder. To initiate and maintain a swarm in place, flying insects often use a visual external cue called a marker. In mosquitoes, understanding the swarming behavior and its relation to the marker has an additional medical relevance since swarming often precedes mating in the wild, thus…
▽ More
Insect swarms are a model system for understanding collective behavior where the collective motion appears in disorder. To initiate and maintain a swarm in place, flying insects often use a visual external cue called a marker. In mosquitoes, understanding the swarming behavior and its relation to the marker has an additional medical relevance since swarming often precedes mating in the wild, thus constituting an important stage to intercept for controlling mosquito population. In this paper, we conduct preliminary experiments to characterize the visual coupling between a swarm of mosquitoes and a marker. A laboratory microcosm with artificial lighting was built to stimulate consistent swarming in the malarial mosquito Anopheles stephensi. The experimental setup was used to film a mosquito swarm with a stereo camera system as a marker was moved back-and-forth with different frequencies. System identification analysis of the frequency response shows that the relationship between the swarm and the marker can be described by delayed second order dynamics in a feedback loop. Further, the length of the internal time delay appears to correlate with the number of mosquitoes swarming on the marker indicating that such a delay may be able capture social interactions within swarming systems. For insect swarms, model fitting of trajectory data provides a way to numerically compare swarming behaviors of different species with respect to marker characteristics. These preliminary results motivate investigating linear dynamic system in feedback as a framework for modeling insect swarms and set the stage for future studies.
△ Less
Submitted 25 August, 2020; v1 submitted 8 July, 2020;
originally announced July 2020.
-
Named Entity Recognition for Nepali Language
Authors:
Oyesh Mann Singh,
Ankur Padia,
Anupam Joshi
Abstract:
Named Entity Recognition have been studied for different languages like English, German, Spanish and many others but no study have focused on Nepali language. In this paper we propose a neural based Nepali NER using latest state-of-the-art architecture based on grapheme-level which doesn't require any hand-crafted features and no data pre-processing. Our novel neural based model gained relative im…
▽ More
Named Entity Recognition have been studied for different languages like English, German, Spanish and many others but no study have focused on Nepali language. In this paper we propose a neural based Nepali NER using latest state-of-the-art architecture based on grapheme-level which doesn't require any hand-crafted features and no data pre-processing. Our novel neural based model gained relative improvement of 33% to 50% compared to feature based SVM model and up to 10% improvement over state-of-the-art neural based model developed for languages beside Nepali.
△ Less
Submitted 15 August, 2019;
originally announced August 2019.
-
Distances between surfaces in 4-manifolds
Authors:
Oliver Singh
Abstract:
If $Σ$ and $Σ'$ are homotopic embedded surfaces in a $4$-manifold then they may be related by a regular homotopy (at the expense of introducing double points) or by a sequence of stabilisations and destabilisations (at the expense of adding genus). This naturally gives rise to two integer-valued notions of distance between the embeddings: the singularity distance $d_{\text{sing}}(Σ,Σ')$ and the st…
▽ More
If $Σ$ and $Σ'$ are homotopic embedded surfaces in a $4$-manifold then they may be related by a regular homotopy (at the expense of introducing double points) or by a sequence of stabilisations and destabilisations (at the expense of adding genus). This naturally gives rise to two integer-valued notions of distance between the embeddings: the singularity distance $d_{\text{sing}}(Σ,Σ')$ and the stabilisation distance $d_{\text{st}}(Σ,Σ')$. Using techniques similar to those used by Gabai in his proof of the 4-dimensional light-bulb theorem, we prove that $d_{\text{st}}(Σ,Σ')\leq d_{\text{sing}}(Σ,Σ')+1$.
△ Less
Submitted 17 February, 2020; v1 submitted 2 May, 2019;
originally announced May 2019.
-
Numerical method based on Galerkin approximation for the fractional advection-dispersion equation
Authors:
Harendra Singh,
Manas Ranjan Sahoo,
Om Prakash Singh
Abstract:
We use a concept of weak asymptotic solution for homogeneous as well as non-homogeneous fractional advection dispersion type equations. Using Legendre scaling functions as basis, a numerical method based on Galerkin approximation is proposed. This leads to a system of fractional ordinary differential equations whose solutions in turn give approximate solution for the advection-dispersion equations…
▽ More
We use a concept of weak asymptotic solution for homogeneous as well as non-homogeneous fractional advection dispersion type equations. Using Legendre scaling functions as basis, a numerical method based on Galerkin approximation is proposed. This leads to a system of fractional ordinary differential equations whose solutions in turn give approximate solution for the advection-dispersion equations of fractional order. Under certain assumptions on the approximate solutions, it is shown that this sequence of approximate solutions forms a weak asymptotic solution. Numerical examples are given to show the effectiveness of the proposed method.
△ Less
Submitted 30 April, 2015;
originally announced April 2015.
-
Automatic Segmentation of Manipuri (Meiteilon) Word into Syllabic Units
Authors:
Kishorjit Nongmeikapam,
Vidya Raj RK,
Oinam Imocha Singh,
Sivaji Bandyopadhyay
Abstract:
The work of automatic segmentation of a Manipuri language (or Meiteilon) word into syllabic units is demonstrated in this paper. This language is a scheduled Indian language of Tibeto-Burman origin, which is also a very highly agglutinative language. This language usages two script: a Bengali script and Meitei Mayek (Script). The present work is based on the second script. An algorithm is designed…
▽ More
The work of automatic segmentation of a Manipuri language (or Meiteilon) word into syllabic units is demonstrated in this paper. This language is a scheduled Indian language of Tibeto-Burman origin, which is also a very highly agglutinative language. This language usages two script: a Bengali script and Meitei Mayek (Script). The present work is based on the second script. An algorithm is designed so as to identify mainly the syllables of Manipuri origin word. The result of the algorithm shows a Recall of 74.77, Precision of 91.21 and F-Score of 82.18 which is a reasonable score with the first attempt of such kind for this language.
△ Less
Submitted 17 July, 2012;
originally announced July 2012.
-
A New Local Adaptive Thresholding Technique in Binarization
Authors:
T. Romen Singh,
Sudipta Roy,
O. Imocha Singh,
Tejmani Sinam,
Kh. Manglem Singh
Abstract:
Image binarization is the process of separation of pixel values into two groups, white as background and black as foreground. Thresholding plays a major in binarization of images. Thresholding can be categorized into global thresholding and local thresholding. In images with uniform contrast distribution of background and foreground like document images, global thresholding is more appropriate. In…
▽ More
Image binarization is the process of separation of pixel values into two groups, white as background and black as foreground. Thresholding plays a major in binarization of images. Thresholding can be categorized into global thresholding and local thresholding. In images with uniform contrast distribution of background and foreground like document images, global thresholding is more appropriate. In degraded document images, where considerable background noise or variation in contrast and illumination exists, there exists many pixels that cannot be easily classified as foreground or background. In such cases, binarization with local thresholding is more appropriate. This paper describes a locally adaptive thresholding technique that removes background by using local mean and mean deviation. Normally the local mean computational time depends on the window size. Our technique uses integral sum image as a prior processing to calculate local mean. It does not involve calculations of standard deviations as in other local adaptive techniques. This along with the fact that calculations of mean is independent of window size speed up the process as compared to other local thresholding techniques.
△ Less
Submitted 25 January, 2012;
originally announced January 2012.
-
The power law character of off-site power failures
Authors:
A. John Arul,
C. Senthil Kumar,
S. Marimuthu,
Om Pal Singh
Abstract:
A study on the behavior of off-site AC power failure recovery times at three nuclear plant sites is presented. It is shown, that power law is appropriate for the representation of failure frequency-duration correlation function of off-site power failure events, based on simple assumptions about component failure and repair rates. It is also found that the annual maxima of power failure duration…
▽ More
A study on the behavior of off-site AC power failure recovery times at three nuclear plant sites is presented. It is shown, that power law is appropriate for the representation of failure frequency-duration correlation function of off-site power failure events, based on simple assumptions about component failure and repair rates. It is also found that the annual maxima of power failure duration follow Frechet distribution, which is a type II asymptotic distribution, strengthening our assumption of power law for the parent distribution. The extreme value distributions obtained are used to extrapolate for failure durations beyond the observed range.
△ Less
Submitted 17 March, 2003;
originally announced March 2003.
-
Orbit Feedback using X-ray Beam Position Monitoring at the Advanced Photon Source
Authors:
Glenn Decker,
Om Singh
Abstract:
The Advanced Photon Source (APS) was commissioned in 1995 as a third-generation x-ray user facility. At that time orbit control was performed exclusively with broadband rf beam position monitors (BPMs). Since then, emphasis has been placed on incorporating x-ray beam position monitors into the orbit control algorithms. This has resulted in an order of magnitude improvement in long-term beam stab…
▽ More
The Advanced Photon Source (APS) was commissioned in 1995 as a third-generation x-ray user facility. At that time orbit control was performed exclusively with broadband rf beam position monitors (BPMs). Since then, emphasis has been placed on incorporating x-ray beam position monitors into the orbit control algorithms. This has resulted in an order of magnitude improvement in long-term beam stability vertically, using x-ray BPMs (X-BPMs) on bending magnet beamlines. Additional processing will allow similar improvements horizontally, once systematic effects associated with variable insertion device (ID) x-ray beams are properly compensated. Progress to date and upgrade plans will be presented, with an emphasis on the details of the required digital signal processing.
△ Less
Submitted 14 December, 2001;
originally announced December 2001.
