教学课程资源库 更多>>

关键词： Computer engineering   Computer science   Artificial intelligence  

摘要： Image forensics protect the authenticity and integrity of digital images. On the contrary, as the countermeasures of digital forensics, anti-forensics is applied to expose the vulnerability of forensics tools. Consequently, forensics researchers could develop forensics tools against possible new attacks. This dissertation investigation demonstrates two image forensics methods based on convolutional neural network (CNN) and two image anti-forensics methods based on generative adversarial network (GAN). Detecting unsharp masking (USM) sharpened image is the first study in this dissertation. A CNN architecture comprises four convolutional layers and a classification module is proposed to discriminate sharpened images and unsharpened images. The results exhibit the superiority of the proposed CNN model over the existing sharpening detection method, i.e., edge perpendicular ternary coding (EPTC). The second study is to detect recolored images. Unlike the conventional binary classifieds, the proposed method based on CNN can be employed for binary classification as well as multiple labels classification. In order to accelerate the training process, the normalization layer is discarded in the proposed CNN. The proposed model can reach detection accuracy over 90% under all circumstances. The detection performance is perfect even when the images are weakly sharpened. To investigate the possible vulnerabilities of sharpening detectors, a GAN model is proposed to behave as an anti-forensics tool. In this study, after adversarial training, the proposed GAN model generates images with the sharpening features. However, these pictures cannot be regarded as sharpened ones. Observed from the experimental results, even the state-of-the-art sharpening detector based on CNN can be also deceived with the images generated by our proposed model. Finally, the fourth study is to investigate whether GAN can be supervised to generate images that can impede forensics detectors from making corre

关键词： Computer science   Biology   Genetics  

摘要： RNA-RNA interactions (RRIs) are essential in many biological processes, including gene transcription, translation, and localization. They play a critical role in diseases such as cancer and Alzheimer’s. An RNA-RNA interaction algorithm uses a dynamic programming algorithm to predict the secondary structure and suffers very high computational time. Its high complexity (Θ(N3M3) in time and Θ(N2M2) in space) makes it both essential and a challenge to parallelize. RRI programs are developed and optimized by hand most of the time, which is prone to human error and costly to develop and maintain. This thesis presents the parallelization of an RRI program - BPMax on a single shared memory CPU platform. From a mathematical specification of the dynamic programming algorithm, we generate highly optimized code that achieves over 100× speedup over the baseline program that uses a standard “diagonal-by-diagonal” execution order. We achieve 100 GFLOPS, which is about a fourth of our platform’s peak theoretical single-precision performance for max-plus computation. The main kernel in the algorithm, whose complexity is Θ(N3M3) attains 186 GFLOPS. We do this with a polyhedral code generation tool, A L P H A Z, which takes user-specified mapping directives and automatically generates optimized C code that enhances parallelism and locality. AL P H AZ allows the user to explore various schedules, memory maps, parallelization approaches, and tiling of the most dominant part of the computation.

关键词： Information technology   Biomedical engineering   Computer science   Artificial intelligence  

摘要： La Dystrophie myotonique de type 1 (DM1) est une maladie neuromusculaire incurable qui, entre autres, cause une perte de capacité musculaire au fil des années. Il a été démontré qu’un programme d’entraînement prescrit par un spécialiste peut être bénéfique pour les personnes atteintes de cette maladie. D’ailleurs, le projet de maîtrise présenté dans ce mémoire se greffe au projet Acti-DM1 ayant pour but le développement d’une solution technologique d’assistance à domicile pour aider ce type de clientèle à réaliser un programme d’entraînement à domicile. La technologie consiste principalement en un bracelet à centrale inertielle enregistrant différentes activités physiques (se lever, s’asseoir, marcher, courir) en temps réel connectée à une unité centrale de traitement possédant un haut-parleur. L’unité centrale de traitement s’occupe de recueillir les données du bracelet et de gérer le programme d’entraînement à domicile prescrit en effectuant des rappels quotidiens et en guidant le participant lors de ses exercices. Au cours du projet Acti-DM1, la technologie d’assistance développée a été déployée dans 10 résidences de personnes atteintes de DM1. Lors de cette cueillette de données en situation réelle d’une durée de 10 semaines, beaucoup de données ont été recueillies. Le projet de recherche dont il est question dans ce mémoire a pour objectif de traiter ces données et de détecter les changements comportementaux au niveau de l’activité physique enregistrée par le bracelet à centrale inertielle au cours de la cueillette. Pour y arriver, des prétraitements importants ont dû être effectués sur les données brutes afin d’extraire l’information utile. Ensuite, plusieurs algorithmes constituant l’état de l’art au niveau de la détection automatique de changements comportementaux ont été étudiés, implémentés et appliqués.

关键词： Libraries   Tools   Systematics   Pediatrics   Internet of Things   Search engines   Presses   Artificial intelligence   Internet of Things   machine learning   ubiquitous computing  

摘要： Violence against women and children is a public health issue of pandemic proportions. It is estimated that one in every three women worldwide has experienced physical, emotional, or sexual violence. Similarly, each year one out of two children are victims of some form of violence including domestic aggression and bullying. Due to the widespread use of the Internet and social media, women and children are now vulnerable to other types of violence such as cyber-bullying and online sexual or emotional harassment. To help alleviate this social problem, the use of computer science and related technologies has been leveraged in recent years. The Internet of Things, artificial intelligence, ubiquitous and mobile computing, pattern recognition, cloud computing and similar technologies, have been used to formulate solutions to detect and prevent violent acts against women and children. In this paper, a systematic review of some of the efforts that can help address the problem of violence against women and children is presented. This paper describes the current state-of-the-art of these contributions and identifies trends, architectures, technologies, and current open challenges. The survey was developed using a literature review of academic documents published from 2010 to 2020. The contributions were categorized into four application domains: online detection, offline detection, safety, and education. These contributions were further categorized based on the computer science approaches and technologies used: artificial intelligence, Internet of Things, and digital serious games.

关键词： Computer science   Design   Artificial intelligence   Computer engineering  

摘要： Modeling and predicting transistors’ behavior with machine learning is a high-dimensional regression problem. It has a scenario that unlabeled data is meaningful, the query has equal cost and features are numeric. During the modeling, a situation that model performs poorly in regions with high variation in output space was observed. To gain a better performance with smaller data size, a query synthesis active learning strategy and a passive learning strategy are proposed to allocate more samples in high-variant regions. The active learning approach is oriented by model performance. In this strategy, the scope would be split into multiple disjoint cells in input space to estimate and optimize their performance individually. By this way, the samples would be mainly taken in poorly performing cells. The passive learning, called Matrix Smooth, is designed based on a different idea: Since the high-variant regions can be located according to model performance, the poorly performed regions for model can also be identified by finding high-variant regions. Thus, this strategy is oriented by data analysis. It supposes the variation of output space can be reflected by the geometric angles between samples in both of input and output space. By stipulating the initial samples are sampled evenly in input space, the features are orthogonal to each other, then the angles would be only influenced by output space. New samples would be iteratively taken near the sharpest angle to smooth the matrix. Compared to active learning, passive learning avoids training model in each sampling iteration, thus significantly reduces the computational cost. The model also takes advantage of determined data size for hyper-parameter tuning. The experiments on Support Vector Regression and Deep Neural Networks show that both proposed strategies outperform evenly sampling. Especially, the Matrix Smooth is found having the ability of avoiding overfitting on Deep Neural Networks.

关键词： Space stations   Algorithms   Surveillance   Space surveillance   Artificial satellites   Aerospace engineering   Computer science   Engineering  

摘要： Space Situational Awareness (SSA) is an important and pressing field, as the amount of space debris is growing due to the increase in spacecraft launches. Space agencies like NASA and ESA keep track of over 5,000 spacecraft and 20,000 spacecraft-related debris objects currently orbiting the Earth. These spacecraft may undergo trajectory changes through artificial impulsive maneuvers without warning, posing a risk to other spacecraft in the vicinity. Additionally, measurement data often consists of uncertainties, where usual spacecraft maneuver detection and characterisation methods are often computationally expensive. The goal of this work is to develop a quick, reliable, computationally inexpensive method for spacecraft maneuver detection and characterisation, which would be used to assist satellite operators in collision avoidance measures. This dissertation utilises such a technique, which is applied to simulated measurement data of an orbiting spacecraft taken from one ground site. With these objectives, a method using a modified Extended Kalman Filter (EKF) with covariance inflation from previous research was explored and simplified. Through building on this set of research, a new parameter definition is presented for maneuver detection and characterisation, named K, with focus on transverse apogee and perigee maneuvers. A relationship was then found between the detection parameter (both the original parameter, Ψ, and the new parameter, K) and the magnitude of the impulsive artificial maneuver. The detection and characterisation method was tested in several different scenarios, varying ∆v magnitude and scale, ∆v direction, maneuver orbital location, orbit geometry, noise covariance, detection parameter threshold, state covariance inflation threshold, and maneuver duration. In general, it was found that as the ∆v magnitude increases, so does the magnitude of the detection parameter. This relationship was found to be nonlinear, of either a quadratic (Ψ) or logari

关键词： Computer science   Artificial intelligence  

摘要： In this thesis, we develop methodologies to enhance the robustness of current neural models for online toxicity detection. Specifically, we aim at adding predictive power and interpretability to transformer-based models. To improve the predictive power of a transformer-based model, we propose to further pre-train the model on the domain-related corpus, i.e., social media text. To add interpretability to a transformer-based model, we introduce a simple and effective assumption, that a post is at least as toxic as its most toxic span, to empower the model with the ability to explain its output during prediction. We incorporate this assumption into transformer-based models by scoring a post based on the maximum toxicity of its spans and augmenting the training process to identify correct spans. The experiments have shown that our proposed idea of further pre-training can improve the model's performance for toxicity detection. We also find our proposed approach that incorporates interpretability does not injure the predictive power of the model and can produce explanations that exceed the quality of those provided by Logistic Regression analysis (often regarded as a highly interpretable model), according to a human study. We also find that our proposed approach can be generalized to different transformer-based models and even different domain tasks.

关键词： Computer science   Artificial intelligence   Men   Neurons   Cameras   Accuracy   Validity   Datasets   Photographs   Collaboration   Autonomous vehicles   Gender   Neural networks   Classification   Shopping   Software engineering   Bias  

摘要： Nowadays, deep neural networks based software have been widely applied in many areas including safety-critical areas such as traffic control, medical diagnosis and malware detection, etc. However, the software engineering techniques, which are supposed to guarantee the functionality, safety as well as fairness, are not well studied. For example, some serious crashes of DNN based autonomous cars have been reported. These crashes could have been avoided if these DNN based software were well tested. Traditional software testing, debugging or repairing techniques do not work well on DNN based software because there is no control flow, data flow or AST(Abstract Syntax Tree) in deep neural networks. Proposing software engineering techniques targeted on DNN based software are imperative. In this thesis, we first introduced the development of SE(Software Engineering) for AI(Artificial Intelligence) area and how our works have influenced the advancement of this new area. Then we summarized related works and some important concepts in SE for AI area. Finally, we discussed four important works of ours. Our first project DeepTest is one of the first few papers proposing systematic software testing techniques for DNN based software. We proposed neuron coverage guided image synthesis techniques for DNN based autonomous cars and leveraged domain specific metamorphic relation to generate oracle for new generated test cases to automatically test DNN based software. We applied DeepTest to testing three top performing self-driving car models in Udacity self-driving car challenge and our tool has identified thousands of erroneous behaviors that may lead to potential fatal crash. In DeepTest project, we found that the natural variation such as spatial transformations or rain/fog effects have led to problematic corner cases for DNN based self-driving cars. In the follow-up project DeepRobust, we studied per-point robustness of deep neural network under natural variation. We found that fo

关键词： Civil engineering   Artificial intelligence   Remote sensing   Computer science  

摘要： In this thesis, a methodology and framework were created to detect the seawalls accurately and efficiently in low coastal areas and was evaluated in the study area of Hallandale Beach City, Broward County, Florida. Aerial images collected from the Florida Department of Transportation (FDOT) were processed using eCognition Developer software for Multi-Resolution Segmentation and Classification of objects. Two classification approaches, pixel-based image analysis, and the object-based image analysis (OBIA) method were applied for image classification. However, Pixel based classification was discarded for having less accuracy in output. Three techniques within object-based classification-machine learning technique, knowledge-based technique and machine learning followed by knowledge-based technique were used to compare the most efficient method of classification. While performing the machine learning technique, three algorithms: Random Forest, support vector machine and decision tree were applied to test the best algorithm. Of all the approaches used, the combination of machine learning and a knowledge-based method was able to map the sea wall effectively.

关键词： Electrical engineering   Artificial intelligence   Computer science  

摘要： Enabling robotic systems for autonomous actions such as driverless systems, is a very complex task in real-world scenarios due to uncertainties. Machine learning capabilities have been quickly making their way into autonomous systems and industrial robotics technology. They found many applications in every sector, including autonomous vehicles, humanoid robots, drones and many more. In this research we will be implementing artificial intelligence in robotic arm to be able to solve a complex balancing control problem from scratch, without any feedback loop and using state of the art deep reinforcement learning algorithm named DQN. The benchmark problem that is considered as case study, is balancing an inverted pendulum upward using a six-degrees freedom robot arm. Very simple form of this problem has been solved recently using machine learning however under this thesis we made a very complex system of inverted pendulum and implemented in Robot Operating System (ROS) which is very realistic simulation environment. We have not only succeeded to control the pendulum but also added turbulences on the learned model to study its robustness. We observed how the initial learned model is unstable at the presence of turbulence and how random turbulences helps the system to transform to a more robust model. We have also used the robust model in different environment and showed how the model adopt itself with the new physical properties. Using orientation sensor on the tip of the inverted pendulum to get angular velocity, simulation in ROS and having inverted pendulum on ball joint are few highlighted novelties in this thesis in compare previous publications.