Artificial Intelligence and Robotics Commons^™

Dependable Machine Intelligence At The Tactical Edge, Archan Misra, Kasthuri Jayarajah, Dulanga Kaveesha Weerakoon Weerakoon Mudiyanselage, Randy Tandriansyah Daratan, Shuochao Yao, Tarek Abdelzaher

Research Collection School Of Computing and Information Systems

The paper describes a vision for dependable application of machine learning-based inferencing on resource-constrained edge devices. The high computational overhead of sophisticated deep learning learning techniques imposes a prohibitive overhead, both in terms of energy consumption and sustainable processing throughput, on such resource-constrained edge devices (e.g., audio or video sensors). To overcome these limitations, we propose a ``cognitive edge" paradigm, whereby (a) an edge device first autonomously uses statistical analysis to identify potential collaborative IoT nodes, and (b) the IoT nodes then perform real-time sharing of various intermediate state to improve their individual execution of machine intelligence tasks. We provide …

A Data-Driven Approach For Modeling Agents, Hamdi Kavak

Computational Modeling & Simulation Engineering Theses & Dissertations

Agents are commonly created on a set of simple rules driven by theories, hypotheses, and assumptions. Such modeling premise has limited use of real-world data and is challenged when modeling real-world systems due to the lack of empirical grounding. Simultaneously, the last decade has witnessed the production and availability of large-scale data from various sensors that carry behavioral signals. These data sources have the potential to change the way we create agent-based models; from simple rules to driven by data. Despite this opportunity, the literature has neglected to offer a modeling approach to generate granular agent behaviors from data, creating …

Development And Initial Evaluation Of A Reinforced Cue Detection Model To Assess Situation Awareness In Commercial Aircraft Cockpits, Aysen K. Taylor

Engineering Management & Systems Engineering Theses & Dissertations

Commercial transport aircraft of today vary greatly from early aircraft with regards to how the aircraft are controlled and the feedback provided from the machine to the human operator. Over time, as avionics systems became more automated, pilots had less direct control over their aircraft. Much research exists in the literature about automation issues, and several major accidents over the last twenty years spurred interest about how to maintain the benefits of automation while improving the overall human-machine interaction as the pilot is considered the last line of defense.

An important reason for maintaining or even improving overall pilot situation …

Variable Speed Limit Control At Sag Curves Through Connected Vehicles: Implications Of Alternative Communications And Sensing Technologies, Reza Vatani Nezafat

Civil & Environmental Engineering Theses & Dissertations

Connected vehicles (CVs) will enable new applications to improve traffic flow. This study’s focus is to investigate how potential implementation of variable speed limit (VSL) through different types of communication and sensing technologies on CVs may improve traffic flow at a sag curve. At sag curves, the gradient changes from negative to positive values which causes a reduction in the roadway capacity and congestion. A VSL algorithm is developed and implemented in a simulation environment for controlling the inflow of vehicles to a sag curve on a freeway to minimize delays and increase throughput. Both vehicle-to-vehicle (V2V) and infrastructure-to-vehicle (I2V) …

Localizing Little Landmarks With Transfer Learning, Sharad Kumar

Dissertations and Theses

Locating a small object in an image -- like a mouse on a computer desk or the door handle of a car -- is an important computer vision problem to solve because in many real life situations a small object may be the first thing that gets operated upon in the image scene. While a significant amount of artificial intelligence and machine learning research has focused on localizing prominent objects in an image, the area of small object detection has remained less explored. In my research I explore the possibility of using context information to localize small objects in an …

Robotic Motion Generation By Using Spatial-Temporal Patterns From Human Demonstrations, Yongqiang Huang

USF Tampa Graduate Theses and Dissertations

Robots excel in manufacturing facilities because the tasks are repetitive and do not change. However, when the tasks change, which happens in almost all tasks that humans perform daily, such as cutting, pouring, and grasping, etc., robots perform much worse. We aim at teaching robots to perform tasks that are subject to change using demonstrations collected from humans, a problem referred to as learning from demonstration (LfD).

LfD consists of two parts: the data of human demonstrations, and the algorithm that extracts knowledge from the data to perform the same motions. Similarly, this thesis is divided into two parts. The …

Imitating Human Responses Via A Dual-Process Model Approach, Matthew A. Grimm

Theses and Dissertations

Human-autonomous system teaming is becoming more prevalent in the Air Force and in society. Often, the concept of a shared mental model is discussed as a means to enhance collaborative work arrangements between a human and an autonomous system. The idea being that when the models are aligned, the team is more productive due to an increase in trust, predictability, and apparent understanding. This research presents the Dual-Process Model using multivariate normal probability density functions (DPM-MN), which is a cognitive architecture algorithm based on the psychological dual-process theory. The dual-process theory proposes a bipartite decision-making process in people. It labels …

Neural Machine Translation, Quinn M. Lanners, Thomas Laurent

Honors Thesis

Neural Machine Translation is the primary algorithm used in industry to perform machine translation. This state-of-the-art algorithm is an application of deep learning in which massive datasets of translated sentences are used to train a model capable of translating between any two languages. The architecture behind neural machine translation is composed of two recurrent neural networks used together in tandem to create an Encoder Decoder structure. Attention mechanisms have recently been developed to further increase the accuracy of these models. In this senior thesis, the various parts of Neural Machine Translation are explored towards the eventual creation of a tutorial …

Conversational Agent: Developing A Model For Intelligent Agents With Transient Emotional States, Angie Dowdell

Theses and Dissertations

The demonstration of aggressive behaviors in the workplace can usually lead to negative professional consequences. However, combat-related professions may involve life and death decision making on an almost daily basis. This is one occupational area in which major duties may invoke or even require aggression. Today’s Army leaders face a different set of challenges than those faced by leaders of the past. Leaders of today must be trained to be both assertive and aggressive at the appropriate times while also displaying empathy. While this is not a new task for an Army leader, it is one that is approached differently …

Distributed Gibbs: A Linear-Space Sampling-Based Dcop Algorithm, Duc Thien Nguyen, William Yeoh, Hoong Chuin Lau, Roie Zivan

Research Collection School Of Computing and Information Systems

Researchers have used distributed constraint optimization problems (DCOPs) to model various multi-agent coordination and resource allocation problems. Very recently, Ottens et al. proposed a promising new approach to solve DCOPs that is based on confidence bounds via their Distributed UCT (DUCT) sampling-based algorithm. Unfortunately, its memory requirement per agent is exponential in the number of agents in the problem, which prohibits it from scaling up to large problems. Thus, in this article, we introduce two new sampling-based DCOP algorithms called Sequential Distributed Gibbs (SD-Gibbs) and Parallel Distributed Gibbs (PD-Gibbs). Both algorithms have memory requirements per agent that is linear in …

A Study Of Face Embedding In Face Recognition, Khanh Duc Le

Master's Theses

Face Recognition has been a long-standing topic in computer vision and pattern recognition field because of its wide and important applications in our daily lives such as surveillance system, access control, and so on. The current modern face recognition model, which keeps only a couple of images per person in the database, can now recognize a face with high accuracy. Moreover, the model does not need to be retrained every time a new person is added to the database.

By using the face dataset from Digital Democracy, the thesis will explore the capability of this model by comparing it with …

Confusion Prediction From Eye-Tracking Data: Experiments With Machine Learning, Joni Salminen, Mridul Nagpal, Haewoon Kwak, Jisun An, Soon-Gyo Jung, Bernard J. Jansen

Research Collection School Of Computing and Information Systems

Predicting user confusion can help improve information presentation on websites, mobile apps, and virtual reality interfaces. One promising information source for such prediction is eye-tracking data about gaze movements on the screen. Coupled with think-aloud records, we explore if user's confusion is correlated with primarily fixation-level features. We find that random forest achieves an accuracy of more than 70% when prediction user confusion using only fixation features. In addition, adding user-level features (age and gender) improves the accuracy to more than 90%. We also find that balancing the classes before training improves performance. We test two balancing algorithms, Synthetic Minority …

Automation Tax Vs Robot-Tax, Vincent Ooi

Research Collection Yong Pung How School Of Law

The positive impact of developments in technology on the economy has historically outweighed the disruptive impact on employment. Society has benefited from the efficiency gains derived from the application of technology in production, while workers displaced by these technologies have largely been successfully retrained and employed in other jobs. However, the pace of development of the “Fourth Industrial Revolution” now presents a risk of mass displacement of human labour, particularly in tasks that are repetitive and menial. The “Fourth Industrial Revolution” is characterised by significant progress in a closely-linked cluster of areas such as robot dexterity, machine learning, processing power, …

Automatic Acquisition Of Annotated Training Corpora For Test-Code Generation, Magdalena Kacmajor, John D. Kelleher

Articles

Open software repositories make large amounts of source code publicly available. Potentially, this source code could be used as training data to develop new, machine learning-based programming tools. For many applications, however, raw code scraped from online repositories does not constitute an adequate training dataset. Building on the recent and rapid improvements in machine translation (MT), one possibly very interesting application is code generation from natural language descriptions. One of the bottlenecks in developing these MT-inspired systems is the acquisition of parallel text-code corpora required for training code-generative models. This paper addresses the problem of automatically synthetizing parallel text-code corpora …

Optimal Design And Implementation Of Fractal Dipole Antenna Based On Arc Of Trigeminal Tree Structure, Xiaohong Zhang, Xiaoxiao Nan, Yixian Yang

Journal of System Simulation

Abstract: A RFID fractal dipole antenna is designed with the arc of the trigeminal tree structure, which is fed by simple microstrip balun to achieve better impedance matching within employing double parallel line structure and by using the dielectric substrate withε_r=4.4 and tan δ=0.035. The results of electromagnetic property simulation analysis of HFSS show that the return loss can reach 36dB, which is much higher than binary arc-shaped fractal antenna which is 16dB; and the impedance matching is very ideal. By physical processing and testing of the antenna, the measured results are in very good agreement with simulation …

Blind Separation Simulation System Of Sound Signals Based On Time-Frequency Analysis Of Short Time Fourier Transformation, Xiaorong Tong

Journal of System Simulation

Abstract: In order to improve the ability of real-time and separation accuracy, a blind separation algorithm based on time-frequency (TF) analysis of short time Fourier transformation is proposed, and the problems of underdetermined sound signal sorting can be solved effectively. The proposed method avoids some of the drawbacks, relaxes the assumption by allowing the sources to be TF-non-disjoint to a certain extent, and the algorithm can be achieved in the condition of TF-non-disjoint. This method was successfully applied to sound signal sorting system; the time domain waves of signals can be got; and the whole signal sorting process can be …

Evaluation Of Flow Line Structure Of Hospital With Spatial Cognitive Agent, Yunfeng Wang

Journal of System Simulation

Abstract: Hospital flow line is the result of the mutual adaptation of its spatial layout and business process, which should be regarded as an independent analysis unit in the spatial layout design. Aiming at the phenomenon that the flow lines have the characteristics of both space and flow and are difficult to analyze, a spatial cognitive process of agents is established by introducing structural elements from the spatial syntax analysis as the knowledge coding language of agents. Evaluation indexes of hospital flow line structure based on computational experiment are proposed and implemented. The framework of flow line analysis combining flow …

Evaluation And Modeling Of Visual Fatigue In 3d Display Based On Ecg, Danli Wang, Xinpan Yang, Yue Kang, Haichen Hu

Journal of System Simulation

Abstract: This study investigates and models the correlation between visual fatigue and ECG data by collecting the ECG data of the participants during watching 3D movies, combining with the subjective scores, the reaction time, and the questionnaires before and after the experiment. ECG data of 12 subjects were analyzed in both time and frequency domain. The results showed that subjective score and HR increased with the duration of viewing, while PNN50 decreased. Linear model of subjective score is established, and the R² of model is 0.930. Subjective questionnaires show significant changes in overall visual fatigue, dry eyes and nausea symptoms …

Centroidal Voronoi Tessellation With Local Optimization, Tianyu Ye, Yiqun Wang, Dongming Yan, Junhai Yong

Journal of System Simulation

Abstract: Centroidal Voronoi tessellation is a special geometric structure, which has many applications in various fields such as geographical information system, signal processing, mesh generation/optimization, visualization and so on. Due to the highly non-convex nature of the CVT energy function, the existing methods for computing CVT have several drawbacks, which always trap into local minima. We propose generation optimization and stochastic optimization schemes for further reducing the CVT energy. Experimental results show that the proposed method improves both quality and efficiency compared to the recent approaches.

Satellite Formation Keeping And Its Stability Analysis Based On Artificial Potential Field Method, Shengqing Yang, Wenyu Ye, Yubin He, Yabin Wan

Journal of System Simulation

Abstract: In this paper, a control method of satellite formation keeping is investigated based on artificial potential field. Since the N-body problem doesn’t have analytical solutions, a leader-follower formation is proposed as a restricted N-body problem. The stable structure has infinite possible pairs of coordinates in the inertial frame, therefore the Jacobian matrix of the corresponding error dynamic system at the equilibrium point is unsolvable. By introducing a transformation of coordinates from inertial frame to a specific reference frame, the stability analysis of error dynamic system is only related to the structure of formation. By the reduction of error …

Structure Optimization Analysis Of Distillation Kettle Based On Elastic-Plastic Mechanics, Xiaodong Hu, Liu Ning, Xiangsong Qi, Jianpeng Wang

Journal of System Simulation

Abstract: Based on the analysis of structural strength by elastic mechanics, there is a large stress singularity due to the discontinuity of the equipment structure, which makes the calculation result not consistent with the actual stress of the equipment. According to the actual design and structure of a distillation reactor, based on the theory of elastic-plastic mechanics, through modeling, meshing, loading and constraint analysis, the stress distribution of distillation kettle is analyzed; and the wall thickness and stiffener structure are optimized. The results show that the calculation results are reliable and the design of wall thickness is smaller than the …

Overview Of Image Segmentation And Registration For Spine Biological Modeling, Juping Gu, Tianyu Cheng, Hua Liang, Jianping Wang, Zhao Jian, Cao Yong

Journal of System Simulation

Abstract: The prevalence of common spinal fractures is high in this day. Spinal image segmentation and registration are the key step and prerequisite of target recognition, biomechanical modeling and finite element analysis. They are also the key technology of noninvasive surgery navigation. Aiming at the problems of spinal image processing, the classical medical image segmentation and registration algorithm is introduced. It also analyzes the defects and prospects of the future development trend of spine image processing. The research results have certain significance for further understanding of spinal medical image processing and promoting the development of rehabilitation therapy for spinal fractures.

Traffic Patterns For Intelligent Vehicles In The Intersection, Wang Jian, Feng Qi, Xiaowei Hu

Journal of System Simulation

Abstract: With the applications of GPS/DR/MM integrated positioning technology and vehicle infrastructure integration technology, three traffic patterns with which intelligent vehicles go through the intersection are put forward. These three traffic patterns include signal lights adaptive traffic pattern, vehicle self-organizing traffic pattern and vehicle infrastructure integration traffic pattern. They can reduce vehicle delay, avoid traffic conflicts, decrease the probability of traffic accidents, improve traffic efficiency and realize the barrier-free accessibility of intelligent vehicles to the intersection. According to the parameters such as vehicles arrival rate, critical velocity and the average vehicle spacing, a clearance movement model and a circular path …

Modeling And Simulation Of Car-Following On Expressway, Yuhan Jia, Jianping Wu, Yiman Du, Mingyu Liu

Journal of System Simulation

Abstract: Aiming at the high-speed following phenomenon on expressways, an improved car-following model is developed in the paper considering variable safety headway distance. The new model is calibrated and validated by the data collected using an instrumented vehicle in Beijing. Stability analysis of the new model is investigated using the control theory method. The numerical simulations are implemented and the result shows good consistency with theoretical study. The new model is integrated into the traffic simulation system of FLOWSIM. Experiment results show that the improved OV model based FLOWSIM can outperform the original system on some specific conditions.

Evaluation Method Of Imminent Battle Situation In Air Combat, Huimin Cao, Anxiang Huang, Lei Xiang

Journal of System Simulation

Abstract: In the state of air combat, the situation change rapidly. Quick and accurate assessment of the current situation and making decisions is the key to the victory of air combat and the preservation of own strength. Based on the specialty of the modern air battle and pelagic operation, this article adds the political factor, the pilot individual actual combat ability, and the formation cooperative tactical task to fight against the single constraint into the evaluation index system. This makes the evaluation more reasonable and comprehensive and closer to the actual combat, so as to provide more accurate basis for …

3d Flight Trajectory Simulation Generator Based On Track Constraints, Shuguang Sun, Cheng Peng

Journal of System Simulation

Abstract: The airborne navigation equipment needs accurate flight trajectory data to test its performance. According to the requirements of the Beidou receivers, considering the constraint of trajectory and flight performance, a flight trajectory generator is designed using iteration like method. The flight trajectory generator can generate the whole trajectory data and the flight parameters of the aircraft based on the actual flight procedures, and display the data and parameters in real time. The whole flight trajectory simulation data from Beijing International Capital Airport to Guangzhou Baiyun International Airport is generated. The result shows that the flight trajectory generator can generate …

Modular Design Method Of Integrated Avionics Simulation System, Mingbo Zhao, Pan Long, Longwei Jiang, Wang Gang, Guangxing Zhao

Journal of System Simulation

Abstract: In the past, there were some problems in the software design for the flight simulation avionics system such as the decentralized structure and the low degree of modularization. Aiming at these problems, a hierarchical, modular integrated avionics system is constructed. The logic part dispersed in each function module is independent to form the integrated logic module. Each task function module and its sub module are responsible for the calculation of the relevant task function algorithm. Modular design is carried out in a deeper level. By dataflow analysis, each module in this design has strong cohesion and weak coupling and …

Rehabilitation Training System Based On Brain Computer Interface, Banghua Yang, Li Bo

Journal of System Simulation

Abstract: Brain-computer interface (BCI) is a new communication and control method. The BCI rehabilitation training system was developed for rehabilitation training of post-stroke patients by using MATLAB and VC++ hybrid programming method. By decoding the EEG signal, the patient's movement intention can be recognized accurately in real time, then the work of the rehabilitation peripheral is triggered to give feedback from the visual, auditory, and tactile. The patient's active rehabilitation is achieved through the system. The system has the characteristics of user-friendly interface, modular algorithm, high online recognition rate; and it can be used as a reference for BCI-based …

Modeling And Simulation Of Inter-City Passenger Travel Service Flow Based On Petri Net, Qipeng Sun, Guo Xin, Ma Fei

Journal of System Simulation

Abstract: On the basis of inter-city passenger travel chain, the inter-city passenger transport service process is studied; the Petri net language is used for establishing and simulating the model of inter-city passenger transport service process based on the Visual Object Net++; and the simulation results are used for verifying the reliability of the established model with the theory of petri net, while the model is transformed into a homogeneous Markov chain with the process. The detailed analysis on the city subway of the inter-city passenger transport service process was carried out. By modeling and simulating, the derived value of …

High-Pass Difference Features Based Image Quality Assessment, Wang Rui, Li Ping, Bin Sheng, Congbin Qiao, Lizhuang Ma, Enhua Wu

Journal of System Simulation

Abstract: Current methods of image quality assessment only can assess the quality of images under the same type of image distortion. In order to fix such weaknesses, this paper is designed based on the image features of natural scene statistics and proposes a new metric method using high-pass filter for detecting features. The approach computes locally the normalized luminance; selects features such as the difference of RGB channels via high-pass filter, image gradient, sharpness, contrast, etc.; and analyzes and gathers features in the metric method trained by logistic regression. Experimental results show that the proposed method can work efficiently under …