Muhammad Saqib is a Postdoctoral Research Fellow, at the School of Computer Science, Faculty of Engineering & IT at UTS. He completed PhD studies at the School of Computer Science. His research interests are in:
- Pedestrian crowd anaysis using computer vision and machine learning
- Object detection and recognition
- Image and video processing
- Data Analytics
- Deep Learning and Convolutional Neural Network Autumn 2019
- Introduction to Data Analytics 2019
- Application development in .NET
- Crowd analysis, Motion flow analysis, Crowd behavior understanding
- Object detection and recognition
- Deep Convolutional Neural Networks
- Visual C#.NET
- Data Analytics
- Computer Vision
- Image and Video Processing
- Python Programming
© 2019 Elsevier Ltd The use of inefficient household appliances and their poor power quality results in energy wastage in residential buildings. These appliances also force the power system to operate at low power factor which results in an ineffective energy utilization. This paper reports the energy consumption pattern of mostly used household appliances individually and collectively over a year. Their power quality parameters are measured through experimentation to calculate the reactive energy consumed by household appliances. This paper also proposes the reactive energy tariffs to enhance the awareness among domestic consumers to make efficient use of household appliances. Currently the reactive power management is being dealt for only the industrial consumer by imposing low power factor penalty. This research estimated that Lahore Electric Supply Company (LESCO) can generate a revenue of almost 150 million US dollars in one year from household consumers by applying three part tariff scheme on reactive energy. By improving the power factor it is estimated that an energy conservation of 1.1 × 109 kWh per annum is also possible. Thus the proposed tariff for reactive energy encourages the domestic consumers to get involved actively in energy conservation while enabling the energy utilities to transfer more active energy to consumers without the expansion of the distribution network.
Hussain, S, Al-Hitmi, M, Khaliq, S, Hussain, A & Saqib, MA 2019, 'Implementation and comparison of particle swarm optimization and genetic algorithm techniques in combined economic emission dispatch of an independent power plant', Energies, vol. 12, no. 11.View/Download from: UTS OPUS or Publisher's site
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license This paper presents the optimization of fuel cost, emission of NOX, COX, and SOX gases caused by the generators in a thermal power plant using penalty factor approach. Practical constraints such as generator limits and power balance were considered. Two contemporary metaheuristic techniques, particle swarm optimization (PSO) and genetic algorithm (GA), have were simultaneously implemented for combined economic emission dispatch (CEED) of an independent power plant (IPP) situated in Pakistan for different load demands. The results are of great significance as the real data of an IPP is used and imply that the performance of PSO is better than that of GA in case of CEED for finding the optimal solution concerning fuel cost, emission, convergence characteristics, and computational time. The novelty of this work is the parallel implementation of PSO and GA techniques in MATLAB environment employed for the same systems. They were then compared in terms of convergence characteristics using 3D plots corresponding to fuel cost and gas emissions. These results are further validated by comparing the performance of both algorithms for CEED on IEEE 30 bus test bed.
Saqib, M, Khan, SD, Sharma, N & Blumenstein, M 2019, 'Crowd Counting in Low-Resolution Crowded Scenes Using Region-Based Deep Convolutional Neural Networks', IEEE Access, vol. 7, pp. 35317-35329.View/Download from: UTS OPUS or Publisher's site
© 2013 IEEE. Crowd counting and density estimation is an important and challenging problem in the visual analysis of the crowd. Most of the existing approaches use regression on density maps for the crowd count from a single image. However, these methods cannot localize individual pedestrian and therefore cannot estimate the actual distribution of pedestrians in the environment. On the other hand, detection-based methods detect and localize pedestrians in the scene, but the performance of these methods degrades when applied in high-density situations. To overcome the limitations of pedestrian detectors, we proposed a motion-guided filter (MGF) that exploits spatial and temporal information between consecutive frames of the video to recover missed detections. Our framework is based on the deep convolution neural network (DCNN) for crowd counting in the low-to-medium density videos. We employ various state-of-the-art network architectures, namely, Visual Geometry Group (VGG16), Zeiler and Fergus (ZF), and VGGM in the framework of a region-based DCNN for detecting pedestrians. After pedestrian detection, the proposed motion guided filter is employed. We evaluate the performance of our approach on three publicly available datasets. The experimental results demonstrate the effectiveness of our approach, which significantly improves the performance of the state-of-the-art detectors.
Saqib, MA, Kashif, SAR & Mujahid, QZ 2014, 'The study of the impacts introduced by a wind farm, having doubly fed induction generators, on a power system', Renewable Energy and Power Quality Journal, vol. 1, no. 12, pp. 161-165.View/Download from: Publisher's site
© 2014, European Association for the Development of Renewable Energy, Environment and Power Quality (EA4EPQ). All rights reserved. The paper discusses the 5th order model of a doubly fed induction generator which is used to study the impacts of a wind farm at the power system. The integration of wind farms poses serious problems on the stability of a power system. The study also illustrates the need for reactive power compensation when a wind generator is to be connected to a power system. A realistic 50 MW wind farm, proposed in the southern Sindh province of Pakistan, is being simulated to illustrate the dynamics of this integration on the national grid. This preliminary study assumes a constant wind speed for the duration of the studied period.
Arif, M, Saqib, M, Basalamah, S & Naeem, A 2012, 'Counting of moving people in the video using neural network system', Life Science Journal, vol. 9, no. 3, pp. 1384-1392.
Automatic counting of people in the crowd using surveillance visual camera is very useful in effective crowd management, security surveillance, and many more applications. In this paper, we have proposed an intelligent framework to automate the process of people counting in the surveillance video. Foreground (moving people) segmentation from the video is done by combination of different foreground estimation techniques. Texture analysis and foreground pixel area for different segmentation techniques are used to extract the useful features. Neural Network is trained on these features and people counting accuracy of more than 96% is achieved on a benchmark video.
Coluccia, A, Saqib, M, Sharma, N, Blumenstein, M, Magoulianitis, V, Ataloglou, D, Dimou, A, Zarpalas, D, Daras, P, Craye, C, Ardjoune, S, Fascista, A, De La Iglesia, D, Mendez, M, Dosil, R, Gonzalez, I, Schumann, A, Sommer, L, Ghenescu, M, Piatrik, T, De Cubber, G, Nalamati, M & Kapoor, A 2019, 'Drone-vs-bird detection challenge at IEEE AVSS2019', 2019 16th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2019.View/Download from: Publisher's site
© 2019 IEEE. This paper presents the second edition of the 'drone-vs-bird' detection challenge, launched within the activities of the 16-th IEEE International Conference on Advanced Video and Signal-based Surveillance (AVSS). The challenge's goal is to detect one or more drones appearing at some point in video sequences where birds may be also present, together with motion in background or foreground. Submitted algorithms should raise an alarm and provide a position estimate only when a drone is present, while not issuing alarms on birds, nor being confused by the rest of the scene. This paper reports on the challenge results on the 2019 dataset, which extends the first edition dataset provided by the SafeShore project with additional footage under different conditions.
Saqib, M, Saleem, MM, Awan, SU & Rehman, MU 2019, 'Design, modeling and parametric analysis of Chevron shaped electrothermal actuator using low cost metalMUMPS fabrication process', 2018 International Conference on Computing, Electronic and Electrical Engineering, ICE Cube 2018.View/Download from: Publisher's site
© 2018 IEEE. High force and large displacement at the low actuation voltage is the primary concern for the micro actuators to be used for microelectromechanical systems (MEMS) based devices. The electrothermal actuators are distinctive in terms of providing large output displacement and force generation at reasonably low actuation voltages. This paper presents the design, modeling and parametric study of the chevron-shaped electrothermal actuator designed by following the design rules of commercially available low cost MetalMUMPS fabrication process. The material properties of electroplated Nickel are used for the finite element method (FEM) based simulations. The behavior of Chevron-shaped electrothermal actuator for the significant parametric variations is also observed. The proposed chevron-shaped electrothermal actuator achieve the output displacement of 57.602 μm along with 313.11 μN force at the low actuation voltage of 0.2 V.
Nalamati, M, Kapoor, A, Saqib, M, Sharma, N & Blumenstein, M 2019, 'Drone detection in long-range surveillance videos', 2019 16th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2019, International Conference on Advanced Video and Signal Based Surveillance, IEEE, Taipei, Taiwan.View/Download from: UTS OPUS or Publisher's site
© 2019 IEEE. The usage of small drones/UAVs has significantly increased recently. Consequently, there is a rising potential of small drones being misused for illegal activities such as terrorism, smuggling of drugs, etc. posing high-security risks. Hence, tracking and surveillance of drones are essential to prevent security breaches. The similarity in the appearance of small drone and birds in complex background makes it challenging to detect drones in surveillance videos. This paper addresses the challenge of detecting small drones in surveillance videos using popular and advanced deep learning-based object detection methods. Different CNN-based architectures such as ResNet-101 and Inception with Faster-RCNN, as well as Single Shot Detector (SSD) model was used for experiments. Due to sparse data available for experiments, pre-trained models were used while training the CNNs using transfer learning. Best results were obtained from experiments using Faster-RCNN with the base architecture of ResNet-101. Experimental analysis on different CNN architectures is presented in the paper, along with the visual analysis of the test dataset.
Saqib, M, Daud Khan, S, Sharma, N, Scully-Power, P, Butcher, P, Colefax, A & Blumenstein, M 2019, 'Real-Time Drone Surveillance and Population Estimation of Marine Animals from Aerial Imagery', International Conference Image and Vision Computing New Zealand.View/Download from: UTS OPUS or Publisher's site
© 2018 IEEE. Video analysis is being rapidly adopted by marine biologists to asses the population and migration of marine animals. Manual analysis of videos by human observers is labor intensive and prone to error. The automatic analysis of videos using state-of-the-art deep learning object detectors provides a cost-effective way for the study of marine animals population and their ecosystem. However, there are many challenges associated with video analysis such as background clutter, illumination, occlusions, and deformation. Due to the high-density of objects in the images and sever occlusion, current state-of-the-art object often results in multiple detections. Therefore, customized Non-Maxima-Suppression is proposed after the detections to suppress false positives which significantly improves the counting and mean average precision of the detections. An end-to-end deep learning framework of Faster-RCNN  was adopted for detections with base architectures of VGG16 , VGGM  and ZF .
Akmal, R, Arif, MS, Saqib, M & Imran, SA 2018, 'Design Considerations for GaN based X-Band, High Power and High Isolation SPDT T/R Switch', 2018 IEEE MTT-S Latin America Microwave Conference, LAMC 2018 - Proceedings.View/Download from: Publisher's site
© 2018 IEEE. This paper presents the design considerations for GaN based high power and high isolation single pole double throw (SPDT) switch for X-band applications. The effects of device periphery, off-state parasitic capacitance, gate control voltage, gate control resistance and a method to improve isolation performance is presented. Additionally, a high power and high isolation SPDT switch is also designed using 0.25 μm GaN HEMT process design kit. The reported switch shows insertion loss less than -1.7 dB, isolation better than -38 dB over the entire X-band and isolation of -63 dB at 10.5 GHz. The input and output return loss are better than -18 dB from 8 to 12 GHz. At the same time, it exhibits P1 dB compression point of 42 dBm. The designed SPDT switch MMIC covers chip area of 1 × 1.95 mm2.
Saqib, M, Saleem, MM, Mazhar, N, Awan, SU & Ur Rehman, M 2018, 'Design and Modeling of Robust Multi Degree of Freedom Micro Gyroscope with Wide Bandwidth', Proceedings of the 21st International Multi Topic Conference, INMIC 2018.View/Download from: Publisher's site
© 2018 IEEE. Micro gyroscopes are among the rapidly aggrandizing field of the micro electro mechanical systems (MEMS), which is procured with its extensive range of applications around the globe. This paper depicts an innovated approach of designing and modeling of a novel brawny multi-mass gyroscope to acquire dynamic motion amplification and spacious bandwidth. The proposed resonator block is accompanied by a series of proof masses coupled to each other where the primary mass defines the driving unit, secondary mass passes on the dynamic energy, tertiary mass is oscillating mass. The main crux of this design is that increasing the number of masses ameliorates the model with respect to amplification by manifolds and enhances the bandwidth. To analyze this structural model, assistance of finite element method (FEM) based analysis is pursued and simulations are carried out for the examination of distinct frequency modes and frequency response of this robust system. The wide bandwidth of over 3 kHz is achieved by this proposed dynamic model.
Das, A, Sengupta, A, Saqib, M, Pal, U & Blumenstein, M 2018, 'More Realistic and Efficient Face-Based Mobile Authentication using CNNs', Proceedings of the International Joint Conference on Neural Networks, International Joint Conference on Neural Networks, IEEE, Rio de Janeiro, Brazil, pp. 1-8.View/Download from: UTS OPUS or Publisher's site
© 2018 IEEE. In this work, we propose a more realistic and efficient facebased mobile authentication technique using CNNs. This paper discusses and explores an inevitable problem of using face images for mobile authentication, taken from varying distances with a front/selfie camera of the mobile phone. Incidentally, once an individual comes towards a certain distance from the camera, the face images get large and appear over-sized. Simultaneously sharp features of some portions of the face, such as forehead, cheek, and chin are changed completely. As a result, the face features change and the impact increases exponentially once the individual crosses a certain distance and gradually approaches towards the front camera. This work proposes a solution (achieving better accuracy and facial features, whereby face images were cropped and aligned around its close bounding box) to mitigate the aforementioned identified gap. The work investigated different frontier face detection and recognition techniques to justify the proposed solution. Among all the employed methods evaluated, CNNs worked best. For a quantitative comparison of the proposed method, manually cropped face images/annotations of the face images along with their close boundary were prepared. In turn, we have developed a database considering the above-mentioned scenario for 40 individuals, which will be publicly available for academic research purposes. The experimental results achieved indicate a successful implementation of the proposed method and the performance of the proposed technique is also found to be superior in comparison to the existing state-of-the-art.
Saqib, M, Daud Khan, S, Sharma, N & Blumenstein, M 2017, 'Extracting descriptive motion information from crowd scenes', International Conference Image and Vision Computing New Zealand, International Conference on Image and Vision Computing New Zealand, IEEE, Christchurch, New Zealand, pp. 1-6.View/Download from: UTS OPUS or Publisher's site
© 2017 IEEE. An important contribution that automated analysis tools can generate for management of pedestrians and crowd safety is the detection of conflicting large pedestrian flows: this kind of movement pattern, in fact, may lead to dangerous situations and potential threats to pedestrian's safety. For this reason, detecting dominant motion patterns and summarizing motion information from the scene are inevitable for crowd management. In this paper, we develop a framework that extracts motion information from the scene by generating point trajectories using particle advection approach. The trajectories obtained are then clustered by using unsupervised hierarchical clustering algorithm, where the similarity is measured by the Longest Common Sub-sequence (LCS) metric. The achieved motions patterns in the scene are summarized and represented by using color-coded arrows, where speeds of the different flows are encoded with colors, the width of an arrow represents the density (number of people belonging to a particular motion pattern) while the arrowhead represents the direction. This novel representation of crowded scene provides a clutter free visualization which helps the crowd managers in understanding the scene. Experimental results show that our method outperforms state-of-the-art methods.
Saqib, M, Khan, SD, Sharma, N & Blumenstein, M 2018, 'Person Head Detection in Multiple Scales Using Deep Convolutional Neural Networks', Proceedings of the International Joint Conference on Neural Networks, International Joint Conference on Neural Networks, IEEE, Rio de Janeiro, Brazil.View/Download from: UTS OPUS or Publisher's site
© 2018 IEEE. Person detection is an important problem in computer vision with many real-world applications. The detection of a person is still a challenging task due to variations in pose, occlusions and lighting conditions. The purpose of this study is to detect human heads in natural scenes acquired from a publicly available dataset of Hollywood movies. In this work, we have used state-of-the-art object detectors based on deep convolutional neural networks. These object detectors include region-based convolutional neural networks using region proposals for detections. Also, object detectors that detect objects in the single-shot by looking at the image only once for detections. We have used transfer learning for fine-tuning the network already trained on a massive amount of data. During the fine-tuning process, the models having high mean Average Precision (mAP) are used for evaluation of the test dataset. Experimental results show that Faster R-CNN  and SSD MultiBox  with VGG16  perform better than YOLO  and also demonstrate significant improvements against several baseline approaches.
Saqib, M, Khan, SD & Blumenstein, M 2016, 'Detecting dominant motion patterns in crowds of pedestrians', Proceedings of SPIE - The International Society for Optical Engineering, International Conference on Graphic and Image Processing, SPIE, Tokyo, Japan.View/Download from: UTS OPUS or Publisher's site
© 2017 SPIE. As the population of the world increases, urbanization generates crowding situations which poses challenges to public safety and security. Manual analysis of crowded situations is a tedious job and usually prone to errors. In this paper, we propose a novel technique of crowd analysis, the aim of which is to detect different dominant motion patterns in real-time videos. A motion field is generated by computing the dense optical flow. The motion field is then divided into blocks. For each block, we adopt an Intra-clustering algorithm for detecting different flows within the block. Later on, we employ Inter-clustering for clustering the flow vectors among different blocks. We evaluate the performance of our approach on different real-time videos. The experimental results show that our proposed method is capable of detecting distinct motion patterns in crowded videos. Moreover, our algorithm outperforms state-of-the-art methods.
Saqib, M, Khan, SD & Blumenstein, M 2016, 'Texture-based feature mining for crowd density estimation: A study', International Conference Image and Vision Computing New Zealand, International Conference on Image and Vision Computing New Zealand, IEEE, Palmerston North, New Zealand.View/Download from: UTS OPUS or Publisher's site
© 2016 IEEE. Texture feature is an important feature descriptor for many image analysis applications. The objectives of this research are to determine distinctive texture features for crowd density estimation and counting. In this paper, we have comprehensively reviewed different texture features and their different possible combinations to evaluate their performance on pedestrian crowds. A two-stage classification and regression based framework have been proposed for performance evaluation of all the texture features for crowd density estimation and counting. According to the framework, input images are divided into blocks and blocks into cells of different sizes, having varying crowd density levels. Due to perspective distortion, people appearing close to the camera contribute more to the feature vector than people far away. Therefore, features extracted are normalized using a perspective normalization map of the scene. At the first stage, image blocks are classified using multi-class SVM into different density level. At the second stage Gaussian Process Regression is used to re gress low-level features to count. Various texture features and their possible combinations are evaluated on publicly available dataset.
Coluccia, A, Ghenescu, M, Piatrik, T, De Cubber, G, Schumann, A, Sommer, L, Klatte, J, Schuchert, T, Beyerer, J, Farhadi, M, Amandi, R, Aker, C, Kalkan, S, Saqib, M, Sharma, N, Makkah, SDK & Blumenstein, M 2017, 'Drone-vs-Bird detection challenge at IEEE AVSS2017', Proceedings of the 2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017, IEEE International Conference on Advanced Video and Signal Based Surveillance, IEEE, Lecce, Italy, pp. 1-6.View/Download from: UTS OPUS or Publisher's site
© 2017 IEEE. Small drones are a rising threat due to their possible misuse for illegal activities, in particular smuggling and terrorism. The project SafeShore, funded by the European Commission under the Horizon 2020 program, has launched the 'drone-vs-bird detection challenge' to address one of the many technical issues arising in this context. The goal is to detect a drone appearing at some point in a video where birds may be also present: the algorithm should raise an alarm and provide a position estimate only when a drone is present, while not issuing alarms on birds. This paper reports on the challenge proposal, evaluation, and results1.
Saqib, M, Daud Khan, S, Sharma, N & Blumenstein, M 2017, 'A study on detecting drones using deep convolutional neural networks', Proceedings of the 2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017, IEEE International Conference on Advanced Video and Signal Based Surveillance, IEEE, Lecce, Italy.View/Download from: UTS OPUS or Publisher's site
© 2017 IEEE. The object detection is a challenging problem in computer vision with various potential real-world applications. The objective of this study is to evaluate the deep learning based object detection techniques for detecting drones. In this paper, we have conducted experiments with different Convolutional Neural Network (CNN) based network architectures namely Zeiler and Fergus (ZF), Visual Geometry Group (VGG16) etc. Due to sparse data available for training, networks are trained with pre-trained models using transfer learning. The snapshot of trained models is saved at regular interval during training. The best models having high mean Average Precision (mAP) for each network architecture are used for evaluation on the test dataset. The experimental results show that VGG16 with Faster R-CNN perform better than other architectures on the training dataset. Visual analysis of the test dataset is also presented.
Latif, K, Jaffar, M, Javaid, N, Saqib, MN, Qasim, U & Khan, ZA 2012, 'Performance analysis of hierarchical routing protocols in wireless sensor networks', Proceedings - 2012 7th International Conference on Broadband, Wireless Computing, Communication and Applications, BWCCA 2012, pp. 620-625.View/Download from: Publisher's site
This work focusses on analyzing the optimization strategies of routing protocols with respect to energy utilization of sensor nodes in Wireless Sensor Network (WSNs). Different routing mechanisms have been proposed to address energy optimization problem in sensor nodes. Clustering mechanism is one of the popular WSNs routing mechanisms. In this paper, we first address energy limitation constraints with respect to maximizing network life time using linear programming formulation technique. To check the efficiency of different clustering scheme against modeled constraints, we select four cluster based routing protocols, Low Energy Adaptive Clustering Hierarchy (LEACH), Threshold Sensitive Energy Efficient sensor Network (TEEN), Stable Election Protocol (SEP), and Distributed Energy Efficient Clustering (DEEC). To validate our mathematical framework, we perform analytical simulations in MATLAB by choosing number of alive nodes, number of dead nodes, number of packets and number of CHs, as performance metrics. © 2012 IEEE.
Kaleem, Z, Lee, CK, Saqib, M, Mohsin, S & Salim, F 2010, 'The way towards amplifier design using CAD (ADS) tool', International Conference on Advanced Communication Technology, ICACT, pp. 962-967.
This paper mainly focuses on design arts and operational mechanism of ADS tool. ADS tool is characterized by reliable, efficient and controlled functioning as compare to conventional approaches. In this paper, we describe an ADS tool based interactive procedure that provides the students in electrical and computer engineering programs with an easy-to-use reference and overview of an amplifier design. This multimedia-based system covers topics that start with introductory basic concepts in amplifier design and conclude with advanced and detailed concepts using the ADS tool.
Saqib, M & Lee, C 2010, 'Traffic control system using wireless sensor network', International Conference on Advanced Communication Technology, ICACT, pp. 352-357.
The Real time locating system (RTLS) determines and tracks the location of assets and people. This paper presents a novel application to estimate the position and velocity of vehicle using wireless sensor network. Two Anchor nodes are used as reader along roadside and total distance between them is known. Whenever a moving vehicle with tag comes in between the common part of the operating range of two anchor nodes, exchange of information is done using Symmetric double sided two way ranging algorithm, which gives us position information. Using position information at several interval of time, velocity can be easily obtained. Position and velocity is obtained and displayed on base station. Kalman filtering is used to estimate the position and velocity from noisy measurements. Performance evaluation is done comparing vehicle position speed true values with experimental and estimated values.
Khan, A, Saqib, M & Kaleem, Z 2009, 'Functional unit level parallelism in RISC architecture', Proceedings of the 6th International Conference on Frontiers of Information Technology, FIT '09.View/Download from: Publisher's site
This paper presents the design and implementation of RISC processor having five stages pipelined architecture. Functional unit parallelism is exploited through the implementation of pipelining in five stages of RISC processor. The hazards which come to life due to parallelism are data, structural, and control hazards .In order to achieve the true benefits of the parallelism through pipelining; these hazards must be properly handled. The data hazards are solved using bypassing in which we forward the required value of the operand to the succeeding instruction. Structural hazards are solved by implementing three port register file so that two operand reading and one register writing can be performed in parallel without degrading the performance. Control hazards arise from Branch, Jump and Call instructions. To solve these problems, we insert automated NOP in stage2, stage3 and stage4. The processor designed is a fully functional processor which can execute any program including jump statements, switch statements, loops and subroutines which are the basic ingredients of any computer program. Copyright 2009 ACM.