International Journal of Advances in Intelligent Informatics
Video-based action quality assessment (AQA) is a non-trivial task due to the subtle visual differ... more Video-based action quality assessment (AQA) is a non-trivial task due to the subtle visual differences between data produced by experts and non-experts. Current methods are extended from the action recognition domain where most are based on temporal pattern matching. AQA has additional requirements where order and tempo matter for rating the quality of an action. We present a novel dataset of ranked TikTok dance videos, and a pairwise AQA method for predicting which video of a same-label pair was sourced from the better dancer. Exhaustive pairings of same-label videos were randomly assigned to 100 human annotators, ultimately producing a ranked list per label category. Our method relies on a successful detection of the subject’s 2D pose inside successive query frames where the order and tempo of actions are encoded inside a produced String sequence. The detected 2D pose returns a top-matching Visual word from a Codebook to represent the current frame. Given a same-label pair, we gen...
10th International Conference on Robotics, Vision, Signal Processing and Power Applications, 2019
Embedded Ethernet and Controller Area Network (CAN) protocol can be used in control network to ac... more Embedded Ethernet and Controller Area Network (CAN) protocol can be used in control network to achieve hard real-time communication. For embedded Ethernet protocol, Carrier Sense Multiple Access with Collision Detection (CSMA/CD) is the media access control (MAC) used to control data transmission between nodes in network. Back-off algorithm in CSMA/CD is used to handle packet collisions and retransmission. For CAN protocol that developed for automotive application, it has priority arbitration to handle collisions and retransmission. In this paper, embedded Ethernet network models and CAN network models are developed and simulated in MATLAB Simulink software. Several back-off algorithms, which are Binary Exponential Backoff (BEB), Linear Back-off Algorithm, Exponential-Linear back-off Algorithm and Logarithm Back-off Algorithm are proposed and implemented into Embedded Ethernet network model to evaluate the performance. Both embedded Ethernet and CAN network models are extended to 3, 10 and 15 nodes to evaluate performance at different network condition. The performance criteria evaluated and discussed are average delay and jitter of packets. The results show that in network with high number of nodes, Linear Back-off Algorithm and Exponential-Linear back-off Algorithm shows improvement in packets delay, 61% and jitter, 83% compared to standard algorithm, BEB. For CAN network, the packet jitter is relatively low, 0.293 ms.
The unprecedented outbreak of novel coronavirus 2019 (COVID-19) globally has a huge impact to our... more The unprecedented outbreak of novel coronavirus 2019 (COVID-19) globally has a huge impact to our daily life in numerous ways. To effectively minimize the spread of the virus, early symptom detection is crucial, especially in closed environment with high human traffic areas which post higher chances of human-to-human transmission. Body temperature measurement has been identified among the vital monitoring parameters. However, current available temperature monitoring mechanism is costly, limited to single individual and limited to locally without integrating to cloud and database. This led to difficulty in effective surveillance for suspicious COVID cases. Hence, the purpose of this paper is to introduce an end-to-end Internet of Things-enabled application for thermal monitoring as an early signal detection and screening method. This work integrates Raspberry Pi, thermal sensor, LCD display, buzzer, and LED light with Raspbian and Restful API for device-to-cloud communication. The sy...
IAES International Journal of Artificial Intelligence (IJ-AI), 2021
Deep reinforcement learning (DRL) which involved reinforcement learning and artificial neural net... more Deep reinforcement learning (DRL) which involved reinforcement learning and artificial neural network allows agents to take the best possible actions to achieve goals. Spiking Neural Network (SNN) faced difficulty in training due to the non-differentiable spike function of spike neuron. In order to overcome the difficulty, Deep Q network (DQN) and Deep Q learning with normalized advantage function (NAF) are proposed to interact with a custom environment. DQN is applied for discrete action space whereas NAF is implemented for continuous action space. The model is trained and tested to validate its performance in order to balance the firing rate of excitatory and inhibitory population of spike neuron by using both algorithms. Training results showed both agents able to explore in the custom environment with OpenAI Gym framework. The trained model for both algorithms capable to balance the firing rate of excitatory and inhibitory of the spike neuron. NAF achieved 0.80% of the average p...
Abstract : Brain Computer Interface (BCI) is a new feature of human-machine interaction for a d... more Abstract : Brain Computer Interface (BCI) is a new feature of human-machine interaction for a direct communication channel from the brain. It involves the extraction of information from brain activity and translates it into system commands using feature extraction and classification algorithms. The study uses signals previously recorded in the BCI lab. Feature selection and classification were based on the Neural Network (NN), Support Vector Machine (SVM) and Linear Discriminant Analysis (LDA). The results of classification show that LDA classifier recorded the highest accuracy in 3 and 4-class of movement compared with SVM and NN classifiers. LDA classified the 4-class of movements at central channel and single channel with the average accuracy of 43.75% and 42%. Further, LDA performed better result in 3-class of movement, with an average accuracy 62%. The highest accuracy for bandpower performed by LDA classifier with average accuracy 41.75 % at beta band.
2020 International Conference on Electronics, Information, and Communication (ICEIC)
The advent of internet enabled small form factor computational devices have already revolutionize... more The advent of internet enabled small form factor computational devices have already revolutionized the way we fetch information, envision intelligent systems by inferencing real-time analytics from these IoT devices. The current trend is focused towards enabling IoT edge gateways as intermediary computational resources contrary to the cloud model that performs all the heavy lifting in the cloud. It is estimated that billions of IoT devices will be deployed by year 2020, however, very little to no information is present on ease of device provisioning. Similarly, the IoT Edge based vertical markets concepts has begun to surface, however, very little information is available that discusses the merits and demerits of this newly envisioned network architecture. This paper aims to implement a Health-care domain scenario by provisioning and connecting real-time device data from simulated as well as emulated virtual IoT devices on industry leading IoT platforms. The results provide a deeper understanding of system performance by evaluating network latencies with increased payloads which further signifies the role and need for deploying Edge IoT gateways within the network. Device provisioning, service profiling and the ease of group resource management is also presented which helps to build larger scalable networks on these IoT platforms.
The old economic and social growth model, characterized by centralized fossil energy consumption,... more The old economic and social growth model, characterized by centralized fossil energy consumption, is progressively shifting, and the third industrial revolution, represented by new energy and Internet technology, is gaining traction. Energy Internet, as a core technology of the third industrial revolution, aims to combine renewable energy and Internet technology to promote the large-scale use and sharing of distributed renewable energy as well as the integration of multiple complex network systems, such as electricity, transportation, and natural gas. This novel technology enables power networks to save energy. However, multienergy synchronization optimization poses a significant problem. As a solution, this study proposed an optimized approach based on the concept of layered control–collaborate optimization. The proposed method allows the distributed device to plan the heat, cold, gas, and electricity in the regional system in the most efficient way possible. Moreover, the proposed...
2020 17th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)
Energy conservation and optimization remains the top researched field for wireless sensor network... more Energy conservation and optimization remains the top researched field for wireless sensor networks, which is one of a subsets and the underlying communication medium for Internet of Things (IoT) devices. These constrained IoT devices are mostly battery operated and therefore requires robust and optimized algorithms to improve resources utilization which inherently increases the life-span for these devices without compromising Quality of Service (QoS). The communication radios on these nodes are the most power hogging components. Therefore, a major focus has always been on MAC and cross-layer protocols to optimize the duty cycle of radios for the conservation of energy. This paper presents a unique scheme for dynamically adjusting the duty cycle of nodes based on the arrival of incoming infrequent source node sensor data over which eliminates the need for frequent periodic channel assessment for network activity. The proposed scheme also makes use of ultra-low wakeUp receivers on the receiver nodes to further aid the node in energy conservation. In this paper, we describe the details of our design scheme, implementation and evaluation details in Contiki OS and Cooja simulator. The results are micro-benchmarked with ContikiMAC and X-MAC protocols, and an improvement in radio duty cycle is reported for lighter network traffic.
Bulletin of Electrical Engineering and Informatics, 2021
The state-of-the-art robust H∞ linear parameter-varying controller is designed for wide speed ope... more The state-of-the-art robust H∞ linear parameter-varying controller is designed for wide speed operating range for non-linear mathematical model of permanent magnet synchronous machines (PMSM) in d-q reference frame for fully electric vehicle. This study propose polytopic approach using rotor speed as scheduling variable to reformulate mathematical model of PMSM into linear parameter varying (LPV) form. The weights were optimized for sensitivity and complementary sensitivity function. The simulation results illustrate fast tracking and enhanced performance of the proposed control technique over wide range of rotor speed. Moreover, as part of this work, the results of H∞ linear parameter varying controller is validated by comparing it with linear quadratic integrator and proportional integral derivative (PID) control techniques to show the effectiveness of the proposed control technique.
The highly efficient Interior Permanent Magnet Synchronous Motor (IPMSM) is ubiquitous choice in ... more The highly efficient Interior Permanent Magnet Synchronous Motor (IPMSM) is ubiquitous choice in Electric Vehicles (EVs) for today’s automotive industry. IPMSM control requires accurate knowledge of an immeasurable critical Permanent Magnet (PM) flux linkage parameter. The PM flux linkage is highly influenced by operating temperature which results in torque derating and hence power loss, unable to meet road loads and reduced life span of electrified powertrain in EVs. In this paper, novel virtual sensing scheme for estimating PM flux linkage through measured stator currents is designed for an IPMSM centric electrified powertrain. The proposed design is based on a Uniform Robust Exact Differentiator (URED) centric Super Twisting Algorithm (STA), which ensures robustness and finite-time convergence of the time derivative of the quadrature axis stator current of IPMSM. Moreover, URED is able to eliminate chattering without sacrificing robustness and precision. The proposed design detec...
This dataset reports the discharge profiles of 4 battery chemistries under IEEE 802.15.4 radio lo... more This dataset reports the discharge profiles of 4 battery chemistries under IEEE 802.15.4 radio load profiles. The batteries were independently subjected to a five-step method to record the discharge characteristics. These discharge currents, their effect on battery capacity, and surface temperature may affect the overall battery lifetime, which was the major aim to discover. The following batteries were used in these experiements <b>Tag</b> <b>Manufacturer</b> <b>Model</b> <b>Battery Chemistry</b> <b>Capacity (mAh)</b> <b>Nominal Voltage</b> <b>C-Rate</b> Batt1 Powerizer MH-AAA1000APZ Nickel-Metal Hydride (Ni-mh) 1000 1.2 V 1C Batt2 Data Power Technology DTP603450 Polymer Lithium-Ion (LiPo) 1000 3.7V 1C Batt3 Panasonic UF553443ZU Lithium-Ion (Li-ion) 1000 3.6V 1C Batt4 Energizer LR-6 Alkaline (Zinc, Magnesium Dioxide) Variable, load dependent 1.5V 2C The five step methodlogy proceeded as: Pre-conditi...
International Journal of Advances in Intelligent Informatics
Video-based action quality assessment (AQA) is a non-trivial task due to the subtle visual differ... more Video-based action quality assessment (AQA) is a non-trivial task due to the subtle visual differences between data produced by experts and non-experts. Current methods are extended from the action recognition domain where most are based on temporal pattern matching. AQA has additional requirements where order and tempo matter for rating the quality of an action. We present a novel dataset of ranked TikTok dance videos, and a pairwise AQA method for predicting which video of a same-label pair was sourced from the better dancer. Exhaustive pairings of same-label videos were randomly assigned to 100 human annotators, ultimately producing a ranked list per label category. Our method relies on a successful detection of the subject’s 2D pose inside successive query frames where the order and tempo of actions are encoded inside a produced String sequence. The detected 2D pose returns a top-matching Visual word from a Codebook to represent the current frame. Given a same-label pair, we gen...
10th International Conference on Robotics, Vision, Signal Processing and Power Applications, 2019
Embedded Ethernet and Controller Area Network (CAN) protocol can be used in control network to ac... more Embedded Ethernet and Controller Area Network (CAN) protocol can be used in control network to achieve hard real-time communication. For embedded Ethernet protocol, Carrier Sense Multiple Access with Collision Detection (CSMA/CD) is the media access control (MAC) used to control data transmission between nodes in network. Back-off algorithm in CSMA/CD is used to handle packet collisions and retransmission. For CAN protocol that developed for automotive application, it has priority arbitration to handle collisions and retransmission. In this paper, embedded Ethernet network models and CAN network models are developed and simulated in MATLAB Simulink software. Several back-off algorithms, which are Binary Exponential Backoff (BEB), Linear Back-off Algorithm, Exponential-Linear back-off Algorithm and Logarithm Back-off Algorithm are proposed and implemented into Embedded Ethernet network model to evaluate the performance. Both embedded Ethernet and CAN network models are extended to 3, 10 and 15 nodes to evaluate performance at different network condition. The performance criteria evaluated and discussed are average delay and jitter of packets. The results show that in network with high number of nodes, Linear Back-off Algorithm and Exponential-Linear back-off Algorithm shows improvement in packets delay, 61% and jitter, 83% compared to standard algorithm, BEB. For CAN network, the packet jitter is relatively low, 0.293 ms.
The unprecedented outbreak of novel coronavirus 2019 (COVID-19) globally has a huge impact to our... more The unprecedented outbreak of novel coronavirus 2019 (COVID-19) globally has a huge impact to our daily life in numerous ways. To effectively minimize the spread of the virus, early symptom detection is crucial, especially in closed environment with high human traffic areas which post higher chances of human-to-human transmission. Body temperature measurement has been identified among the vital monitoring parameters. However, current available temperature monitoring mechanism is costly, limited to single individual and limited to locally without integrating to cloud and database. This led to difficulty in effective surveillance for suspicious COVID cases. Hence, the purpose of this paper is to introduce an end-to-end Internet of Things-enabled application for thermal monitoring as an early signal detection and screening method. This work integrates Raspberry Pi, thermal sensor, LCD display, buzzer, and LED light with Raspbian and Restful API for device-to-cloud communication. The sy...
IAES International Journal of Artificial Intelligence (IJ-AI), 2021
Deep reinforcement learning (DRL) which involved reinforcement learning and artificial neural net... more Deep reinforcement learning (DRL) which involved reinforcement learning and artificial neural network allows agents to take the best possible actions to achieve goals. Spiking Neural Network (SNN) faced difficulty in training due to the non-differentiable spike function of spike neuron. In order to overcome the difficulty, Deep Q network (DQN) and Deep Q learning with normalized advantage function (NAF) are proposed to interact with a custom environment. DQN is applied for discrete action space whereas NAF is implemented for continuous action space. The model is trained and tested to validate its performance in order to balance the firing rate of excitatory and inhibitory population of spike neuron by using both algorithms. Training results showed both agents able to explore in the custom environment with OpenAI Gym framework. The trained model for both algorithms capable to balance the firing rate of excitatory and inhibitory of the spike neuron. NAF achieved 0.80% of the average p...
Abstract : Brain Computer Interface (BCI) is a new feature of human-machine interaction for a d... more Abstract : Brain Computer Interface (BCI) is a new feature of human-machine interaction for a direct communication channel from the brain. It involves the extraction of information from brain activity and translates it into system commands using feature extraction and classification algorithms. The study uses signals previously recorded in the BCI lab. Feature selection and classification were based on the Neural Network (NN), Support Vector Machine (SVM) and Linear Discriminant Analysis (LDA). The results of classification show that LDA classifier recorded the highest accuracy in 3 and 4-class of movement compared with SVM and NN classifiers. LDA classified the 4-class of movements at central channel and single channel with the average accuracy of 43.75% and 42%. Further, LDA performed better result in 3-class of movement, with an average accuracy 62%. The highest accuracy for bandpower performed by LDA classifier with average accuracy 41.75 % at beta band.
2020 International Conference on Electronics, Information, and Communication (ICEIC)
The advent of internet enabled small form factor computational devices have already revolutionize... more The advent of internet enabled small form factor computational devices have already revolutionized the way we fetch information, envision intelligent systems by inferencing real-time analytics from these IoT devices. The current trend is focused towards enabling IoT edge gateways as intermediary computational resources contrary to the cloud model that performs all the heavy lifting in the cloud. It is estimated that billions of IoT devices will be deployed by year 2020, however, very little to no information is present on ease of device provisioning. Similarly, the IoT Edge based vertical markets concepts has begun to surface, however, very little information is available that discusses the merits and demerits of this newly envisioned network architecture. This paper aims to implement a Health-care domain scenario by provisioning and connecting real-time device data from simulated as well as emulated virtual IoT devices on industry leading IoT platforms. The results provide a deeper understanding of system performance by evaluating network latencies with increased payloads which further signifies the role and need for deploying Edge IoT gateways within the network. Device provisioning, service profiling and the ease of group resource management is also presented which helps to build larger scalable networks on these IoT platforms.
The old economic and social growth model, characterized by centralized fossil energy consumption,... more The old economic and social growth model, characterized by centralized fossil energy consumption, is progressively shifting, and the third industrial revolution, represented by new energy and Internet technology, is gaining traction. Energy Internet, as a core technology of the third industrial revolution, aims to combine renewable energy and Internet technology to promote the large-scale use and sharing of distributed renewable energy as well as the integration of multiple complex network systems, such as electricity, transportation, and natural gas. This novel technology enables power networks to save energy. However, multienergy synchronization optimization poses a significant problem. As a solution, this study proposed an optimized approach based on the concept of layered control–collaborate optimization. The proposed method allows the distributed device to plan the heat, cold, gas, and electricity in the regional system in the most efficient way possible. Moreover, the proposed...
2020 17th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)
Energy conservation and optimization remains the top researched field for wireless sensor network... more Energy conservation and optimization remains the top researched field for wireless sensor networks, which is one of a subsets and the underlying communication medium for Internet of Things (IoT) devices. These constrained IoT devices are mostly battery operated and therefore requires robust and optimized algorithms to improve resources utilization which inherently increases the life-span for these devices without compromising Quality of Service (QoS). The communication radios on these nodes are the most power hogging components. Therefore, a major focus has always been on MAC and cross-layer protocols to optimize the duty cycle of radios for the conservation of energy. This paper presents a unique scheme for dynamically adjusting the duty cycle of nodes based on the arrival of incoming infrequent source node sensor data over which eliminates the need for frequent periodic channel assessment for network activity. The proposed scheme also makes use of ultra-low wakeUp receivers on the receiver nodes to further aid the node in energy conservation. In this paper, we describe the details of our design scheme, implementation and evaluation details in Contiki OS and Cooja simulator. The results are micro-benchmarked with ContikiMAC and X-MAC protocols, and an improvement in radio duty cycle is reported for lighter network traffic.
Bulletin of Electrical Engineering and Informatics, 2021
The state-of-the-art robust H∞ linear parameter-varying controller is designed for wide speed ope... more The state-of-the-art robust H∞ linear parameter-varying controller is designed for wide speed operating range for non-linear mathematical model of permanent magnet synchronous machines (PMSM) in d-q reference frame for fully electric vehicle. This study propose polytopic approach using rotor speed as scheduling variable to reformulate mathematical model of PMSM into linear parameter varying (LPV) form. The weights were optimized for sensitivity and complementary sensitivity function. The simulation results illustrate fast tracking and enhanced performance of the proposed control technique over wide range of rotor speed. Moreover, as part of this work, the results of H∞ linear parameter varying controller is validated by comparing it with linear quadratic integrator and proportional integral derivative (PID) control techniques to show the effectiveness of the proposed control technique.
The highly efficient Interior Permanent Magnet Synchronous Motor (IPMSM) is ubiquitous choice in ... more The highly efficient Interior Permanent Magnet Synchronous Motor (IPMSM) is ubiquitous choice in Electric Vehicles (EVs) for today’s automotive industry. IPMSM control requires accurate knowledge of an immeasurable critical Permanent Magnet (PM) flux linkage parameter. The PM flux linkage is highly influenced by operating temperature which results in torque derating and hence power loss, unable to meet road loads and reduced life span of electrified powertrain in EVs. In this paper, novel virtual sensing scheme for estimating PM flux linkage through measured stator currents is designed for an IPMSM centric electrified powertrain. The proposed design is based on a Uniform Robust Exact Differentiator (URED) centric Super Twisting Algorithm (STA), which ensures robustness and finite-time convergence of the time derivative of the quadrature axis stator current of IPMSM. Moreover, URED is able to eliminate chattering without sacrificing robustness and precision. The proposed design detec...
This dataset reports the discharge profiles of 4 battery chemistries under IEEE 802.15.4 radio lo... more This dataset reports the discharge profiles of 4 battery chemistries under IEEE 802.15.4 radio load profiles. The batteries were independently subjected to a five-step method to record the discharge characteristics. These discharge currents, their effect on battery capacity, and surface temperature may affect the overall battery lifetime, which was the major aim to discover. The following batteries were used in these experiements <b>Tag</b> <b>Manufacturer</b> <b>Model</b> <b>Battery Chemistry</b> <b>Capacity (mAh)</b> <b>Nominal Voltage</b> <b>C-Rate</b> Batt1 Powerizer MH-AAA1000APZ Nickel-Metal Hydride (Ni-mh) 1000 1.2 V 1C Batt2 Data Power Technology DTP603450 Polymer Lithium-Ion (LiPo) 1000 3.7V 1C Batt3 Panasonic UF553443ZU Lithium-Ion (Li-ion) 1000 3.6V 1C Batt4 Energizer LR-6 Alkaline (Zinc, Magnesium Dioxide) Variable, load dependent 1.5V 2C The five step methodlogy proceeded as: Pre-conditi...
Uploads
Papers