Muhammad Umar Farooq - Academia.edu (original) (raw)

Papers by Muhammad Umar Farooq

arXiv (Cornell University), Feb 13, 2020

Braided convolutional codes (BCCs) are a class of spatially coupled turbo-like codes that can be ... more Braided convolutional codes (BCCs) are a class of spatially coupled turbo-like codes that can be described by a (2, 3)-regular compact graph. In this paper, we introduce a family of (dv, dc)-regular GLDPC codes with convolutional code constraints (CC-GLDPC codes), which form an extension of classical BCCs to arbitrary regular graphs. In order to characterize the performance in the waterfall and error floor regions, we perform an analysis of the density evolution thresholds as well as the finitelength ensemble weight enumerators and minimum distances of the ensembles. In particular, we consider various ensembles of overall rate R = 1/3 and R = 1/2 and study the trade-off between variable node degree and strength of the component codes. We also compare the results to corresponding classical LDPC codes with equal degrees and rates. It is observed that for the considered LDPC codes with variable node degree dv > 2, we can find a CC-GLDPC code with smaller dv that offers similar or better performance in terms of BP and MAP thresholds at the expense of a negligible loss in the minimum distance.

Processes, Aug 8, 2022

This article is an open access article distributed under the terms and conditions of the Creative... more This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY

International journal of endorsing health science research, Dec 1, 2014

Objective: To evaluate salivary detection of interleukin 6 & 8 and high risk HPV-16 & 18 are info... more Objective: To evaluate salivary detection of interleukin 6 & 8 and high risk HPV-16 & 18 are informative biomarkers of Oral Pre-malignant Lesion (PML) and Oral Squamous Cell Carcinoma (OSCC) in our population. Duration: July 2011 to December 2012. Subjects and Methods: Total 105 cases were included. The subjects were divided in three groups 'A', 'B' & 'C' having 35 participants each. Group 'A' comprised of patients having strong clinical evidence of oral PML. Group 'B' constitutes histologically proven OSCC and Group 'C' includes disease free subjects as controls. Relevant clinical history was recorded after informed consent on institutional approved performa. Saliva was collected as per standard drooling method'. Samples were stored at +4oC and later transferred to Dow Diagnostic, Research & Reference Laboratory to store it at-20oC before further process. Samples were subjected to centrifugation at 4500 rpm for 15 minutes at 4oC. Supernatant fluid phase was used in ELISA for detection and quantification of IL6 and IL8.. Cell pellets were used for identification of high risk HPV-16 & 18 by real-time PCR. Data was entered and analyzed on SPSS version 16. P-value of 0.05 was taken as standard reference. Results: In group 'A', IL6 was not detected in almost all the subjects except one case. IL8 was detected in 26/35 (74.3%) subjects and not detected in 09 (25.7%) cases. In group 'B', IL6 was detected in 13 (37.1%) cases and in 22 (62.9%) cases, it cannot be detected. IL8 was detected in 33 (94.3%) and it was not detected in 02 (5.7%) subjects. It is observed that IL8 is consistently found raised in group 'A' & 'B'. In group 'C', IL6 was not detected in any of the subject while IL8 was detected in 10(28.6%) cases. Significant association was found for qualitative salivary detection of IL6 and IL8 between the groups (P= < 0.0001 and < 0.0001 respectively). Regarding quantitative salivary concentration of IL6 & IL8, no significant co-relation was found in salivary levels of IL6 between the groups while there was significant association of salivary IL8 levels between the groups (P= <0.0001). On post Hoc multiple comparison, significant co-relation was found in IL8 levels between oral PML group and controls (P=0.001) and OSCC group and controls (P= <0.0001). In group 'A', HPV-16 was detected in salivary samples of 3 (8.6%) cases while HPV-18 was not detected. In group 'B', HPV-16 was detected in the salivary samples of 07 (20%) cases while HPV-18 was detected in 06 (17.1%) cases. Mixed HPV-16 and HPV-18 were found in 02 (5.7%) cases. In group 'C', HPV-16 was detected in 03(8.6%) cases while HPV-18 was not detected in any of the subjects. Significant relationship was observed between the groups for salivary HPV-18 detection (P= 0.002) while for detection of HPV-16, no significant association was found (P= 0.245). Conclusion: Salivary concentration of IL6 and IL8 in oral PML and oral cancer are useful biomarkers in our population. Detection of HPV infection for the causation of oral cancer cannot be fully established possibly due to small sample size. More over different genetic makeup, environmental and geographic differences, indulgence in peculiar risk factor habits and different sexual practices compared to west due to socio-cultural and religious restrictions could be the reason.

2022 International Telecommunications Conference (ITC-Egypt), Jul 26, 2022

Ambient computing is getting popular as one of the most substantial technological advances in the... more Ambient computing is getting popular as one of the most substantial technological advances in the future. In the present era, human activity tracking, indoor localization, and healthcare systems are all developing rapidly. Researchers are able to find practical solutions in healthcare facilities that often need to locate humans with the growing affordability and power of Radio Frequency (RF) technology. RF is appealing to monitor human activities in an unobtrusive and remote manner. Channel State Information (CSI) can be used as a contactless method to identify and locate human activity indoors. This paper presents the results of an experiment utilizing Universal Software-Defined Radio Peripherals (USRP) to locate the location of activity. A single subject is observed performing sitting, standing, no activity and leaning forward in six different locations inside a room to collect CSI samples. Additional CSI is collected when the subject walks in both directions within the designated area. Three Machine Learning (ML) classification algorithms were used in the comparison: Random Forest, Extra Trees (ET), and Multilayer Perceptron (MLP). When compared to other ML algorithms, the ET classifier has the best performance, with an average of 95% accuracy.

List of Tables xiii List of Figures xiv Chapter 1. Introduction

IEEE Access

In this paper, a novel rotation and scale invariant approach for texture classification based on ... more In this paper, a novel rotation and scale invariant approach for texture classification based on Gabor filters has been proposed. These filters are designed to capture the visual content of the images based on their impulse responses which are sensitive to rotation and scaling in the images. The filter responses are rearranged according to the filter exhibiting the response having largest amplitude, followed by the calculation of patterns after binarizing the responses based on a particular threshold. This threshold is obtained as the average energy of Gabor filter responses at a particular pixel. The binary patterns are converted to decimal numbers, the histograms of which are used as texture features. The proposed features are used to classify the images from two famous texture datasets: Brodatz, CUReT and UMD texture albums. Experiments show that the proposed feature extraction method performs really well when compared with several other state-of-the-art methods considered in this paper and is more robust to noise.

arXiv (Cornell University), Jul 7, 2022

This is a repository copy of Non-linear pairwise language mappings for low-resource multilingual ... more This is a repository copy of Non-linear pairwise language mappings for low-resource multilingual acoustic model fusion.

arXiv (Cornell University), Nov 30, 2022

CC-GLPDC codes are a class of generalized lowdensity parity-check (GLDPC) codes where the constra... more CC-GLPDC codes are a class of generalized lowdensity parity-check (GLDPC) codes where the constraint nodes (CNs) represent convolutional codes. This allows for efficient decoding in the trellis with the forward-backward algorithm, and the strength of the component codes easily can be controlled by the encoder memory without changing the graph structure. In this letter, we extend the class of CC-GLDPC codes by introducing different types of irregularity at the CNs and investigating their effect on the BP and MAP decoding thresholds for the binary erasure channel (BEC). For the considered class of codes, an exhaustive grid search is performed to find the BP-optimized and MAP-optimized ensembles and compare their thresholds with the regular ensemble of the same design rate. The results show that irregularity can significantly improve the BP thresholds, whereas the thresholds of the MAP-optimized ensembles are only slightly different from the regular ensembles. Simulation results for the AWGN channel are presented as well and compared to the corresponding thresholds.

Lecture notes in networks and systems, 2022

Humans are able to perceive objects only in the visible spectrum range which limits the perceptio... more Humans are able to perceive objects only in the visible spectrum range which limits the perception abilities in poor weather or low illumination conditions. The limitations are usually handled through technological advancements in thermographic imaging. However, thermal cameras have poor spatial resolutions compared to RGB cameras. Super-resolution (SR) techniques are commonly used to improve the overall quality of low-resolution images. There has been a major shift of research among the Computer Vision researchers towards SR techniques particularly aimed for thermal images. This paper analyzes the performance of three deep learning-based state-of-the-art SR algorithms namely Enhanced Deep Super Resolution (EDSR), Residual Channel Attention Network (RCAN) and Residual Dense Network (RDN) on thermal images. The algorithms were trained from scratch for different upscaling factors of ×2 and ×4. The dataset was generated from two different thermal imaging sequences of BU-TIV benchmark. The sequences contain both sparse and highly dense type of crowds with a far field camera view. The trained models were then used to super-resolve unseen test images. The quantitative analysis of the test images was performed using common image quality metrics such as PSNR, SSIM and LPIPS, while qualitative analysis was provided by evaluating effectiveness of the algorithms for crowd counting application. After only 54 and 51 epochs of RCAN and RDN respectively, both approaches were able to output average scores of 37.878, 0.986, 0.0098 and 30.175, 0.945, 0.0636 for PSNR, SSIM and LPIPS respectively. The EDSR algorithm took the least computation time during both training and testing because of its simple architecture. This research proves that a reasonable accuracy can be achieved with fewer training epochs when an application-specific dataset is carefully selected.

Interspeech 2022

Multilingual speech recognition has drawn significant attention as an effective way to compensate... more Multilingual speech recognition has drawn significant attention as an effective way to compensate data scarcity for lowresource languages. End-to-end (e2e) modelling is preferred over conventional hybrid systems, mainly because of no lexicon requirement. However, hybrid DNN-HMMs still outperform e2e models in limited data scenarios. Furthermore, the problem of manual lexicon creation has been alleviated by publicly available trained models of grapheme-to-phoneme (G2P) and text to IPA transliteration for a lot of languages. In this paper, a novel approach of hybrid DNN-HMM acoustic models fusion is proposed in a multilingual setup for the low-resource languages. Posterior distributions from different monolingual acoustic models, against a target language speech signal, are fused together. A separate regression neural network is trained for each source-target language pair to transform posteriors from source acoustic model to the target language. These networks require very limited data as compared to the ASR training. Posterior fusion yields a relative gain of 14.65% and 6.5% when compared with multilingual and monolingual baselines respectively. Cross-lingual model fusion shows that the comparable results can be achieved without using posteriors from the language dependent ASR.

Journal of the Brazilian Society of Mechanical Sciences and Engineering

The structural integrity of additive manufacturing structures is a pronounced challenge consideri... more The structural integrity of additive manufacturing structures is a pronounced challenge considering the voids and weak layer-to-layer adhesion. One of the potential ways is hybrid deposition manufacturing (HDM) that includes fused filament fabrication (FFF) with the conventional filling process, also known as “HDM composites". HDM is a potential technique for improving structural stability by replacing the thermoplastic void structure with a voidless epoxy. However, the literature lacks investigation of FFF/epoxy HDM-based composites regarding optimal volume distribution, effects of brittle and ductile FFF materials, and fractographic analysis. This research presents the effects of range of volume distributions (10–90%) between FFF and epoxy system for tensile, flexure, and compressive characterization. Volume distribution in tensile and flexure samples is achieved using printable wall thickness, slot width, and maximum width. For compression, the printable wall thickness, slot...

IEEE Access

The revolution of IoT highly impacts on different applications such as remote sensing, smart citi... more The revolution of IoT highly impacts on different applications such as remote sensing, smart cities, and remote digital healthcare. People use IoT devices for performing business transactions, daily tasks, and healthcare monitoring. IoT devices generate huge amounts of data assets that have potential applications. Biometrics is a potential application of sensors data. The traditional biometric methods such as PINs, passwords are exposed to numerous attacks such as replication, repeated passwords, etc. Sensors' data-based continuous authentication methods are suitable for maintaining users' privacy and security in mobile IoT systems. Most of the existing authentication methods have applied motion-based sensors for building users' identification profiles. The proposed method uses motion sensors and biomedical sensors for reliable and multi-factor user authentication. In this article, we have introduced an IoT sensors data analytics framework to construct user authentication models. We apply the fiducial points-based feature extraction method data for extracting discriminative features. These features act as unique user profiles for authentication purposes. We have performed a detailed analysis of the proposed approach using the publically available datasets. The experiments elaborate on the effectiveness of IoTauth for improved authentication results.

Production Planning & Control, 2022

Highlights o This research aims to investigate the use of lean transformation approach for achiev... more Highlights o This research aims to investigate the use of lean transformation approach for achieving resilience in disaster relief operations (DROs). o A systematic lean based method instigated through DROs' management initiative was developed. o This was validated through an empirical industrial fire case study where selected lean concepts and tools like SIPOC-analysis, value stream maps (VSM), key performance indicators, fish-bone diagram, and plan-do-check-act were used to investigate DROs resilience in terms of their responsiveness. o The VSMs were developed for 'as-is' and 'to-be' scenarios, and comparative analysis against standardized key performance indicators was carried out. o The lean transformation approach was found effective in the studied case of industrial fire for developing resilience in DROs. o Furthermore, lean tools could help in devising pragmatic strategies to prevent delays and achieve higher resilience through better coordination, communication, capacity building and awareness. o This research contributes to the operations management and disaster management fields through lean transformation.

Entropy, 2021

In this paper, we perform a belief propagation (BP) decoding threshold analysis of spatially coup... more In this paper, we perform a belief propagation (BP) decoding threshold analysis of spatially coupled (SC) turbo-like codes (TCs) (SC-TCs) on the additive white Gaussian noise (AWGN) channel. We review Monte-Carlo density evolution (MC-DE) and efficient prediction methods, which determine the BP thresholds of SC-TCs over the AWGN channel. We demonstrate that instead of performing time-consuming MC-DE computations, the BP threshold of SC-TCs over the AWGN channel can be predicted very efficiently from their binary erasure channel (BEC) thresholds. From threshold results, we conjecture that the similarity of MC-DE and predicted thresholds is related to the threshold saturation capability as well as capacity-approaching maximum a posteriori (MAP) performance of an SC-TC ensemble.

The agricultural sector in Pakistan is confronted with the major challenge of significantly incre... more The agricultural sector in Pakistan is confronted with the major challenge of significantly increasing crop productivity to feed a fast-growing population. Decreasing availability and quality of natural resources is adversely affecting crop productivity in the short and long run. Pakistan is facing a severe shortage of water, making the country the fourth most water-stressed country in the world. Providing right amount of water at right time to any crop is important for the development of that crop. This is possible by intervening technology and automating the irrigation systems to be adopted to specific crop needs. The manual irrigation system leads to wastage of water and damaging of the crops due to uninformed watering, while commercially available smart irrigation systems are too expensive to be afforded by farmers. To overcome these problem, a low-cost smart crop monitoring and irrigation system is proposed based on IoT and mobile application. The primary aim of this project is...

PLOS ONE, 2020

In the process of software development, regression testing is one of the major activities that is... more In the process of software development, regression testing is one of the major activities that is done after making modifications in the current system or whenever a software system evolves. But, the test suite size increases with the addition of new test cases and it becomes in-efficient because of the occurrence of redundant, broken, and obsolete test cases. For that reason, it results in additional time and budget to run all these test cases. Many researchers have proposed computational intelligence and conventional approaches for dealing with this problem and they have achieved an optimized test suite by selecting, minimizing or reducing, and prioritizing test cases. Currently, most of these optimization approaches are single objective and static in nature. But, it is mandatory to use multi-objective dynamic approaches for optimization due to the advancements in information technology and associated market challenges. Therefore, we have proposed three variants of self-tunable Ad...

2021 IEEE 93rd Vehicular Technology Conference (VTC2021-Spring), 2021

Spatially coupled serially concatenated codes (SC-SCCs) are a class of spatially coupled turbo-li... more Spatially coupled serially concatenated codes (SC-SCCs) are a class of spatially coupled turbo-like codes, which have a close-to-capacity performance and low error floor. In this paper we investigate the impact of coupling memory, block length, decoding window size, and number of iterations on the performance, complexity, and latency of SC-SCCs. Several design tradeoffs are presented to see the relation between these parameters in a wide range. Also, our analysis provides design guidelines for SC-SCCs in different scenarios to make the code design independent of block length. As a result, block length and coupling memory can be exchanged flexibly without changing the latency and complexity. Also, we observe that the performance of SC-SCCs is improved with respect to the uncoupled ensembles for a fixed latency and complexity.

2020 IEEE International Symposium on Information Theory (ISIT), 2020

Braided convolutional codes (BCCs) are a class of spatially coupled turbo-like codes that can be ... more Braided convolutional codes (BCCs) are a class of spatially coupled turbo-like codes that can be described by a (2, 3)-regular compact graph. In this paper, we introduce a family of (dv, dc)-regular GLDPC codes with convolutional code constraints (CC-GLDPC codes), which form an extension of classical BCCs to arbitrary regular graphs. In order to characterize the performance in the waterfall and error floor regions, we perform an analysis of the density evolution thresholds as well as the finitelength ensemble weight enumerators and minimum distances of the ensembles. In particular, we consider various ensembles of overall rate R = 1/3 and R = 1/2 and study the trade-off between variable node degree and strength of the component codes. We also compare the results to corresponding classical LDPC codes with equal degrees and rates. It is observed that for the considered LDPC codes with variable node degree dv > 2, we can find a CC-GLDPC code with smaller dv that offers similar or better performance in terms of BP and MAP thresholds at the expense of a negligible loss in the minimum distance.

Sustainability, 2021

Pandemics cause chaotic situations in supply chains (SC) around the globe, which can lead towards... more Pandemics cause chaotic situations in supply chains (SC) around the globe, which can lead towards survivability challenges. The ongoing COVID-19 pandemic is an unprecedented humanitarian crisis that has severely affected global business dynamics. Similar vulnerabilities have been caused by other outbreaks in the past. In these terms, prevention strategies against propagating disruptions require vigilant goal conceptualization and roadmaps. In this respect, there is a need to explore supply chain operation management strategies to overcome the challenges that emerge due to COVID-19-like situations. Therefore, this review is aimed at exploring such challenges and developing strategies for sustainability, and viability perspectives for SCs, through a structured literature review (SLR) approach. Moreover, this study investigated the impacts of previous epidemic outbreaks on SCs, to identify the research objectives, methodological approaches, and implications for SCs. The study also expl...

Metals, 2021

Recently, DC53 die steel was introduced to the die and mold industry because of its excellent cha... more Recently, DC53 die steel was introduced to the die and mold industry because of its excellent characteristics i.e., very good machinability and better engineering properties. DC53 demonstrates a strong capability to retain a near-net shape profile of the die, which is a very challenging process with materials. To produce complex and accurate die features, the use of the wire electric discharge machining (WEDM) process takes the lead in the manufacturing industry. However, the challenge is to understand the physical science of the process to improve surface features and service properties. In this study, a detailed yet systematic evaluation of process parameters investigation is made on the influence of a wire feed, pulse on duration, open voltage, and servo voltage on the productivity (material removal rate) and material quality (surface roughness, recast layer thickness, kerf width) against the requirements of mechanical-tooling industry. Based on parametric exploration, wire feed ...

arXiv (Cornell University), Feb 13, 2020

Braided convolutional codes (BCCs) are a class of spatially coupled turbo-like codes that can be ... more Braided convolutional codes (BCCs) are a class of spatially coupled turbo-like codes that can be described by a (2, 3)-regular compact graph. In this paper, we introduce a family of (dv, dc)-regular GLDPC codes with convolutional code constraints (CC-GLDPC codes), which form an extension of classical BCCs to arbitrary regular graphs. In order to characterize the performance in the waterfall and error floor regions, we perform an analysis of the density evolution thresholds as well as the finitelength ensemble weight enumerators and minimum distances of the ensembles. In particular, we consider various ensembles of overall rate R = 1/3 and R = 1/2 and study the trade-off between variable node degree and strength of the component codes. We also compare the results to corresponding classical LDPC codes with equal degrees and rates. It is observed that for the considered LDPC codes with variable node degree dv > 2, we can find a CC-GLDPC code with smaller dv that offers similar or better performance in terms of BP and MAP thresholds at the expense of a negligible loss in the minimum distance.

Processes, Aug 8, 2022

This article is an open access article distributed under the terms and conditions of the Creative... more This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY

International journal of endorsing health science research, Dec 1, 2014

Objective: To evaluate salivary detection of interleukin 6 & 8 and high risk HPV-16 & 18 are info... more Objective: To evaluate salivary detection of interleukin 6 & 8 and high risk HPV-16 & 18 are informative biomarkers of Oral Pre-malignant Lesion (PML) and Oral Squamous Cell Carcinoma (OSCC) in our population. Duration: July 2011 to December 2012. Subjects and Methods: Total 105 cases were included. The subjects were divided in three groups 'A', 'B' & 'C' having 35 participants each. Group 'A' comprised of patients having strong clinical evidence of oral PML. Group 'B' constitutes histologically proven OSCC and Group 'C' includes disease free subjects as controls. Relevant clinical history was recorded after informed consent on institutional approved performa. Saliva was collected as per standard drooling method'. Samples were stored at +4oC and later transferred to Dow Diagnostic, Research & Reference Laboratory to store it at-20oC before further process. Samples were subjected to centrifugation at 4500 rpm for 15 minutes at 4oC. Supernatant fluid phase was used in ELISA for detection and quantification of IL6 and IL8.. Cell pellets were used for identification of high risk HPV-16 & 18 by real-time PCR. Data was entered and analyzed on SPSS version 16. P-value of 0.05 was taken as standard reference. Results: In group 'A', IL6 was not detected in almost all the subjects except one case. IL8 was detected in 26/35 (74.3%) subjects and not detected in 09 (25.7%) cases. In group 'B', IL6 was detected in 13 (37.1%) cases and in 22 (62.9%) cases, it cannot be detected. IL8 was detected in 33 (94.3%) and it was not detected in 02 (5.7%) subjects. It is observed that IL8 is consistently found raised in group 'A' & 'B'. In group 'C', IL6 was not detected in any of the subject while IL8 was detected in 10(28.6%) cases. Significant association was found for qualitative salivary detection of IL6 and IL8 between the groups (P= < 0.0001 and < 0.0001 respectively). Regarding quantitative salivary concentration of IL6 & IL8, no significant co-relation was found in salivary levels of IL6 between the groups while there was significant association of salivary IL8 levels between the groups (P= <0.0001). On post Hoc multiple comparison, significant co-relation was found in IL8 levels between oral PML group and controls (P=0.001) and OSCC group and controls (P= <0.0001). In group 'A', HPV-16 was detected in salivary samples of 3 (8.6%) cases while HPV-18 was not detected. In group 'B', HPV-16 was detected in the salivary samples of 07 (20%) cases while HPV-18 was detected in 06 (17.1%) cases. Mixed HPV-16 and HPV-18 were found in 02 (5.7%) cases. In group 'C', HPV-16 was detected in 03(8.6%) cases while HPV-18 was not detected in any of the subjects. Significant relationship was observed between the groups for salivary HPV-18 detection (P= 0.002) while for detection of HPV-16, no significant association was found (P= 0.245). Conclusion: Salivary concentration of IL6 and IL8 in oral PML and oral cancer are useful biomarkers in our population. Detection of HPV infection for the causation of oral cancer cannot be fully established possibly due to small sample size. More over different genetic makeup, environmental and geographic differences, indulgence in peculiar risk factor habits and different sexual practices compared to west due to socio-cultural and religious restrictions could be the reason.

2022 International Telecommunications Conference (ITC-Egypt), Jul 26, 2022

Ambient computing is getting popular as one of the most substantial technological advances in the... more Ambient computing is getting popular as one of the most substantial technological advances in the future. In the present era, human activity tracking, indoor localization, and healthcare systems are all developing rapidly. Researchers are able to find practical solutions in healthcare facilities that often need to locate humans with the growing affordability and power of Radio Frequency (RF) technology. RF is appealing to monitor human activities in an unobtrusive and remote manner. Channel State Information (CSI) can be used as a contactless method to identify and locate human activity indoors. This paper presents the results of an experiment utilizing Universal Software-Defined Radio Peripherals (USRP) to locate the location of activity. A single subject is observed performing sitting, standing, no activity and leaning forward in six different locations inside a room to collect CSI samples. Additional CSI is collected when the subject walks in both directions within the designated area. Three Machine Learning (ML) classification algorithms were used in the comparison: Random Forest, Extra Trees (ET), and Multilayer Perceptron (MLP). When compared to other ML algorithms, the ET classifier has the best performance, with an average of 95% accuracy.

List of Tables xiii List of Figures xiv Chapter 1. Introduction

IEEE Access

In this paper, a novel rotation and scale invariant approach for texture classification based on ... more In this paper, a novel rotation and scale invariant approach for texture classification based on Gabor filters has been proposed. These filters are designed to capture the visual content of the images based on their impulse responses which are sensitive to rotation and scaling in the images. The filter responses are rearranged according to the filter exhibiting the response having largest amplitude, followed by the calculation of patterns after binarizing the responses based on a particular threshold. This threshold is obtained as the average energy of Gabor filter responses at a particular pixel. The binary patterns are converted to decimal numbers, the histograms of which are used as texture features. The proposed features are used to classify the images from two famous texture datasets: Brodatz, CUReT and UMD texture albums. Experiments show that the proposed feature extraction method performs really well when compared with several other state-of-the-art methods considered in this paper and is more robust to noise.

arXiv (Cornell University), Jul 7, 2022

This is a repository copy of Non-linear pairwise language mappings for low-resource multilingual ... more This is a repository copy of Non-linear pairwise language mappings for low-resource multilingual acoustic model fusion.

arXiv (Cornell University), Nov 30, 2022

CC-GLPDC codes are a class of generalized lowdensity parity-check (GLDPC) codes where the constra... more CC-GLPDC codes are a class of generalized lowdensity parity-check (GLDPC) codes where the constraint nodes (CNs) represent convolutional codes. This allows for efficient decoding in the trellis with the forward-backward algorithm, and the strength of the component codes easily can be controlled by the encoder memory without changing the graph structure. In this letter, we extend the class of CC-GLDPC codes by introducing different types of irregularity at the CNs and investigating their effect on the BP and MAP decoding thresholds for the binary erasure channel (BEC). For the considered class of codes, an exhaustive grid search is performed to find the BP-optimized and MAP-optimized ensembles and compare their thresholds with the regular ensemble of the same design rate. The results show that irregularity can significantly improve the BP thresholds, whereas the thresholds of the MAP-optimized ensembles are only slightly different from the regular ensembles. Simulation results for the AWGN channel are presented as well and compared to the corresponding thresholds.

Lecture notes in networks and systems, 2022

Humans are able to perceive objects only in the visible spectrum range which limits the perceptio... more Humans are able to perceive objects only in the visible spectrum range which limits the perception abilities in poor weather or low illumination conditions. The limitations are usually handled through technological advancements in thermographic imaging. However, thermal cameras have poor spatial resolutions compared to RGB cameras. Super-resolution (SR) techniques are commonly used to improve the overall quality of low-resolution images. There has been a major shift of research among the Computer Vision researchers towards SR techniques particularly aimed for thermal images. This paper analyzes the performance of three deep learning-based state-of-the-art SR algorithms namely Enhanced Deep Super Resolution (EDSR), Residual Channel Attention Network (RCAN) and Residual Dense Network (RDN) on thermal images. The algorithms were trained from scratch for different upscaling factors of ×2 and ×4. The dataset was generated from two different thermal imaging sequences of BU-TIV benchmark. The sequences contain both sparse and highly dense type of crowds with a far field camera view. The trained models were then used to super-resolve unseen test images. The quantitative analysis of the test images was performed using common image quality metrics such as PSNR, SSIM and LPIPS, while qualitative analysis was provided by evaluating effectiveness of the algorithms for crowd counting application. After only 54 and 51 epochs of RCAN and RDN respectively, both approaches were able to output average scores of 37.878, 0.986, 0.0098 and 30.175, 0.945, 0.0636 for PSNR, SSIM and LPIPS respectively. The EDSR algorithm took the least computation time during both training and testing because of its simple architecture. This research proves that a reasonable accuracy can be achieved with fewer training epochs when an application-specific dataset is carefully selected.

Interspeech 2022

Multilingual speech recognition has drawn significant attention as an effective way to compensate... more Multilingual speech recognition has drawn significant attention as an effective way to compensate data scarcity for lowresource languages. End-to-end (e2e) modelling is preferred over conventional hybrid systems, mainly because of no lexicon requirement. However, hybrid DNN-HMMs still outperform e2e models in limited data scenarios. Furthermore, the problem of manual lexicon creation has been alleviated by publicly available trained models of grapheme-to-phoneme (G2P) and text to IPA transliteration for a lot of languages. In this paper, a novel approach of hybrid DNN-HMM acoustic models fusion is proposed in a multilingual setup for the low-resource languages. Posterior distributions from different monolingual acoustic models, against a target language speech signal, are fused together. A separate regression neural network is trained for each source-target language pair to transform posteriors from source acoustic model to the target language. These networks require very limited data as compared to the ASR training. Posterior fusion yields a relative gain of 14.65% and 6.5% when compared with multilingual and monolingual baselines respectively. Cross-lingual model fusion shows that the comparable results can be achieved without using posteriors from the language dependent ASR.

Journal of the Brazilian Society of Mechanical Sciences and Engineering

The structural integrity of additive manufacturing structures is a pronounced challenge consideri... more The structural integrity of additive manufacturing structures is a pronounced challenge considering the voids and weak layer-to-layer adhesion. One of the potential ways is hybrid deposition manufacturing (HDM) that includes fused filament fabrication (FFF) with the conventional filling process, also known as “HDM composites". HDM is a potential technique for improving structural stability by replacing the thermoplastic void structure with a voidless epoxy. However, the literature lacks investigation of FFF/epoxy HDM-based composites regarding optimal volume distribution, effects of brittle and ductile FFF materials, and fractographic analysis. This research presents the effects of range of volume distributions (10–90%) between FFF and epoxy system for tensile, flexure, and compressive characterization. Volume distribution in tensile and flexure samples is achieved using printable wall thickness, slot width, and maximum width. For compression, the printable wall thickness, slot...

IEEE Access

The revolution of IoT highly impacts on different applications such as remote sensing, smart citi... more The revolution of IoT highly impacts on different applications such as remote sensing, smart cities, and remote digital healthcare. People use IoT devices for performing business transactions, daily tasks, and healthcare monitoring. IoT devices generate huge amounts of data assets that have potential applications. Biometrics is a potential application of sensors data. The traditional biometric methods such as PINs, passwords are exposed to numerous attacks such as replication, repeated passwords, etc. Sensors' data-based continuous authentication methods are suitable for maintaining users' privacy and security in mobile IoT systems. Most of the existing authentication methods have applied motion-based sensors for building users' identification profiles. The proposed method uses motion sensors and biomedical sensors for reliable and multi-factor user authentication. In this article, we have introduced an IoT sensors data analytics framework to construct user authentication models. We apply the fiducial points-based feature extraction method data for extracting discriminative features. These features act as unique user profiles for authentication purposes. We have performed a detailed analysis of the proposed approach using the publically available datasets. The experiments elaborate on the effectiveness of IoTauth for improved authentication results.

Production Planning & Control, 2022

Highlights o This research aims to investigate the use of lean transformation approach for achiev... more Highlights o This research aims to investigate the use of lean transformation approach for achieving resilience in disaster relief operations (DROs). o A systematic lean based method instigated through DROs' management initiative was developed. o This was validated through an empirical industrial fire case study where selected lean concepts and tools like SIPOC-analysis, value stream maps (VSM), key performance indicators, fish-bone diagram, and plan-do-check-act were used to investigate DROs resilience in terms of their responsiveness. o The VSMs were developed for 'as-is' and 'to-be' scenarios, and comparative analysis against standardized key performance indicators was carried out. o The lean transformation approach was found effective in the studied case of industrial fire for developing resilience in DROs. o Furthermore, lean tools could help in devising pragmatic strategies to prevent delays and achieve higher resilience through better coordination, communication, capacity building and awareness. o This research contributes to the operations management and disaster management fields through lean transformation.

Entropy, 2021

In this paper, we perform a belief propagation (BP) decoding threshold analysis of spatially coup... more In this paper, we perform a belief propagation (BP) decoding threshold analysis of spatially coupled (SC) turbo-like codes (TCs) (SC-TCs) on the additive white Gaussian noise (AWGN) channel. We review Monte-Carlo density evolution (MC-DE) and efficient prediction methods, which determine the BP thresholds of SC-TCs over the AWGN channel. We demonstrate that instead of performing time-consuming MC-DE computations, the BP threshold of SC-TCs over the AWGN channel can be predicted very efficiently from their binary erasure channel (BEC) thresholds. From threshold results, we conjecture that the similarity of MC-DE and predicted thresholds is related to the threshold saturation capability as well as capacity-approaching maximum a posteriori (MAP) performance of an SC-TC ensemble.

The agricultural sector in Pakistan is confronted with the major challenge of significantly incre... more The agricultural sector in Pakistan is confronted with the major challenge of significantly increasing crop productivity to feed a fast-growing population. Decreasing availability and quality of natural resources is adversely affecting crop productivity in the short and long run. Pakistan is facing a severe shortage of water, making the country the fourth most water-stressed country in the world. Providing right amount of water at right time to any crop is important for the development of that crop. This is possible by intervening technology and automating the irrigation systems to be adopted to specific crop needs. The manual irrigation system leads to wastage of water and damaging of the crops due to uninformed watering, while commercially available smart irrigation systems are too expensive to be afforded by farmers. To overcome these problem, a low-cost smart crop monitoring and irrigation system is proposed based on IoT and mobile application. The primary aim of this project is...

PLOS ONE, 2020

In the process of software development, regression testing is one of the major activities that is... more In the process of software development, regression testing is one of the major activities that is done after making modifications in the current system or whenever a software system evolves. But, the test suite size increases with the addition of new test cases and it becomes in-efficient because of the occurrence of redundant, broken, and obsolete test cases. For that reason, it results in additional time and budget to run all these test cases. Many researchers have proposed computational intelligence and conventional approaches for dealing with this problem and they have achieved an optimized test suite by selecting, minimizing or reducing, and prioritizing test cases. Currently, most of these optimization approaches are single objective and static in nature. But, it is mandatory to use multi-objective dynamic approaches for optimization due to the advancements in information technology and associated market challenges. Therefore, we have proposed three variants of self-tunable Ad...

2021 IEEE 93rd Vehicular Technology Conference (VTC2021-Spring), 2021

Spatially coupled serially concatenated codes (SC-SCCs) are a class of spatially coupled turbo-li... more Spatially coupled serially concatenated codes (SC-SCCs) are a class of spatially coupled turbo-like codes, which have a close-to-capacity performance and low error floor. In this paper we investigate the impact of coupling memory, block length, decoding window size, and number of iterations on the performance, complexity, and latency of SC-SCCs. Several design tradeoffs are presented to see the relation between these parameters in a wide range. Also, our analysis provides design guidelines for SC-SCCs in different scenarios to make the code design independent of block length. As a result, block length and coupling memory can be exchanged flexibly without changing the latency and complexity. Also, we observe that the performance of SC-SCCs is improved with respect to the uncoupled ensembles for a fixed latency and complexity.

2020 IEEE International Symposium on Information Theory (ISIT), 2020

Braided convolutional codes (BCCs) are a class of spatially coupled turbo-like codes that can be ... more Braided convolutional codes (BCCs) are a class of spatially coupled turbo-like codes that can be described by a (2, 3)-regular compact graph. In this paper, we introduce a family of (dv, dc)-regular GLDPC codes with convolutional code constraints (CC-GLDPC codes), which form an extension of classical BCCs to arbitrary regular graphs. In order to characterize the performance in the waterfall and error floor regions, we perform an analysis of the density evolution thresholds as well as the finitelength ensemble weight enumerators and minimum distances of the ensembles. In particular, we consider various ensembles of overall rate R = 1/3 and R = 1/2 and study the trade-off between variable node degree and strength of the component codes. We also compare the results to corresponding classical LDPC codes with equal degrees and rates. It is observed that for the considered LDPC codes with variable node degree dv > 2, we can find a CC-GLDPC code with smaller dv that offers similar or better performance in terms of BP and MAP thresholds at the expense of a negligible loss in the minimum distance.

Sustainability, 2021

Pandemics cause chaotic situations in supply chains (SC) around the globe, which can lead towards... more Pandemics cause chaotic situations in supply chains (SC) around the globe, which can lead towards survivability challenges. The ongoing COVID-19 pandemic is an unprecedented humanitarian crisis that has severely affected global business dynamics. Similar vulnerabilities have been caused by other outbreaks in the past. In these terms, prevention strategies against propagating disruptions require vigilant goal conceptualization and roadmaps. In this respect, there is a need to explore supply chain operation management strategies to overcome the challenges that emerge due to COVID-19-like situations. Therefore, this review is aimed at exploring such challenges and developing strategies for sustainability, and viability perspectives for SCs, through a structured literature review (SLR) approach. Moreover, this study investigated the impacts of previous epidemic outbreaks on SCs, to identify the research objectives, methodological approaches, and implications for SCs. The study also expl...

Metals, 2021

Recently, DC53 die steel was introduced to the die and mold industry because of its excellent cha... more Recently, DC53 die steel was introduced to the die and mold industry because of its excellent characteristics i.e., very good machinability and better engineering properties. DC53 demonstrates a strong capability to retain a near-net shape profile of the die, which is a very challenging process with materials. To produce complex and accurate die features, the use of the wire electric discharge machining (WEDM) process takes the lead in the manufacturing industry. However, the challenge is to understand the physical science of the process to improve surface features and service properties. In this study, a detailed yet systematic evaluation of process parameters investigation is made on the influence of a wire feed, pulse on duration, open voltage, and servo voltage on the productivity (material removal rate) and material quality (surface roughness, recast layer thickness, kerf width) against the requirements of mechanical-tooling industry. Based on parametric exploration, wire feed ...