Computer Sciences Commons^™

Facial Re-Enactment, Speech Synthesis And The Rise Of The Deepfake, Nicholas Gardiner

Theses : Honours

Emergent technologies in the fields of audio speech synthesis and video facial manipulation have the potential to drastically impact our societal patterns of multimedia consumption. At a time when social media and internet culture is plagued by misinformation, propaganda and “fake news”, their latent misuse represents a possible looming threat to fragile systems of information sharing and social democratic discourse. It has thus become increasingly recognised in both academic and mainstream journalism that the ramifications of these tools must be examined to determine what they are and how their widespread availability can be managed.

This research project seeks to examine …

An Automated Snick Detection And Classification Scheme As A Cricket Decision Review System, Aftab Khan, Syed Qadir Hussain, Muhammad Waleed, Ashfaq Khan, Umair Khan

Turkish Journal of Electrical Engineering and Computer Sciences

Umpire decisions can greatly affect the outcome of a cricket game. When there is doubt about the umpire?s call for a decision, a decision review system (DRS) may be brought into play by a batsman or bowler to validate the decision. Recently, the latest technologies, including Hotspot, Hawk-eye, and Snickometer, have been employed when there is doubt among the on-field umpire, batsman, or bowlers. This research is a step forward in gaging the true class of a snick generated from the contact of the cricket ball with either (i) the bat, (ii) gloves, (iii) pad, or (iv) a combination of …

An Explainable Autoencoder For Collaborative Filtering Recommendation, Pegah Sagheb Haghighi, Olurotimi Seton, Olfa Nasraoui

Faculty and Staff Scholarship

Autoencoders are a common building block of Deep Learning architectures, where they are mainly used for representation learning. They have also been successfully used in Collaborative Filtering (CF) recommender systems to predict missing ratings. Unfortunately, like all black box machine learning models, they are unable to explain their outputs. Hence, while predictions from an Autoencoderbased recommender system might be accurate, it might not be clear to the user why a recommendation was generated. In this work, we design an explainable recommendation system using an Autoencoder model whose predictions can be explained using the neighborhood based explanation style. Our preliminary work …

Global Inference For Aspect And Opinion Terms Co-Extraction Based On Multi-Task Neural Networks, Jianfei Yu, Jing Jiang, Rui Xia

Research Collection School Of Computing and Information Systems

Extracting aspect terms and opinion terms are two fundamental tasks in opinion mining. The recent success of deep learning has inspired various neural network architectures, which have been shown to achieve highly competitive performance in these two tasks. However, most existing methods fail to explicitly consider the syntactic relations among aspect terms and opinion terms, which may lead to the inconsistencies between the model predictions and the syntactic constraints. To this end, we first apply a multi-task learning framework to implicitly capture the relations between the two tasks, and then propose a global inference method by explicitly modelling several syntactic …

Automatic Slide Generation For Scientific Papers, Athar Sefid, Jian Wu, Prasenjit Mitra, C. Lee Giles

Computer Science Faculty Publications

We describe our approach for automatically generating presentation slides for scientific papers using deep neural networks. Such slides can help authors have a starting point for their slide generation process. Extractive summarization techniques are applied to rank and select important sentences from the original document. Previous work identified important sentences based only on a limited number of features that were extracted from the position and structure of sentences in the paper. Our method extends previous work by (1) extracting a more comprehensive list of surface features, (2) considering semantic or meaning of the sentence, and (3) using context around the …

Deepsign: A Deep-Learning Architecture For Sign Language, Jai Amrish Shah

Computer Science and Engineering Theses - Archive

Sign languages are used by deaf people for communication. In sign languages, humans use hand gestures, body, facial expressions and movements to convey meaning. Humans can easily learn and understand sign languages, but automatic sign language recognition for machines is a challenging task. Using recent advances in the field of deep-learning, we introduce a fully automated deep-learning architecture for isolated sign language recognition. Our architecture tries to address three problems: 1) Satisfactory accuracy with limited data samples 2) Reducing chances of over-fitting when the data is limited 3) Automating recognition of isolated signs. Our architecture uses deep convolutional encoder-decoder architecture …

Improving Ultra-Wideband Localization By Detecting Radio Misclassification, Cory A. Mayer

Master's Theses

The Global Positioning System (GPS) and other satellite-based positioning systems are often a key component in applications requiring localization. However, accurate positioning in areas with poor GPS coverage, such as inside buildings and in dense cities, is in increasing demand for many modern applications. Fortunately, recent developments in ultra-wideband (UWB) radio technology have enabled precise positioning in places where it was not previously possible by utilizing multipath-resistant wide band pulses.

Although ultra-wideband signals are less prone to multipath interference, it is still a bottleneck as increasingly ambitious projects continue to demand higher precision. Some UWB radios include on-board detection of …

Entity-Grounded Image Captioning, Annika Lindh, Robert J. Ross, John D. Kelleher

Conference papers

An urgent limitation in current Image Captioning models is their tendency to produce generic captions that avoid the interesting detail which makes each image unique. To address this limitation, we propose an approach that enforces a stronger alignment between image regions and specific segments of text. The model architecture is composed of a visual region proposer, a region-order planner and a region-guided caption generator. The region-guided caption generator incorporates a novel information gate which allows visual and textual input of different frequencies and dimensionalities in a Recurrent Neural Network.

Fake News Detection: A Deep Learning Approach, Aswini Thota, Priyanka Tilak, Simrat Ahluwalia, Nibrat Lohia

SMU Data Science Review

Fake news is defined as a made-up story with an intention to deceive or to mislead. In this paper we present the solution to the task of fake news detection by using Deep Learning architectures. Gartner research [1] predicts that “By 2022, most people in mature economies will consume more false information than true information”. The exponential increase in production and distribution of inaccurate news presents an immediate need for automatically tagging and detecting such twisted news articles. However, automated detection of fake news is a hard task to accomplish as it requires the model to understand nuances in natural …

Hyper-Optimized Machine Learning And Deep Learning Methods For Geo-Spatial And Temporal Function Estimation, Neelabh Pant

Computer Science and Engineering Dissertations - Archive

Owing to a high degree of freedom in human mobility, accurate modelling/estimation of human mobility function remains a challenge. Numerous work in the literature have tried to address the challenge using various traditional machine learning methods on spatio-temporal attributes of data. We compare the use of Varied-K Means clustering, Hidden Markov Model techniques, feed forward neural networks, recurrent neural networks (RNN) and Long Short Term Recurrent Neural Networks (LSTM) to predict a user's future movement based on the user's past historical data. Although several techniques were proposed to predict a user's movement, not many have concentrated on a user's location …

Hierarchical Representation Learning With Connectionist Models, De Wang

Computer Science and Engineering Dissertations - Archive

To unleash the power of big data, efficient algorithms which are scalable to millions of data are desired. Deep learning is one area that benefits from big data enormously. Deep learning uses neural networks to mimic human brains, this approach is termed connectionist in AI community. In this dissertation, we propose several novel learning strategies to improve the performance of connectionist models. Evaluation of a large neural network during inference phase requires a lot of GPU memory and computation, which will degrade user experience due to response latency. Model distillation is one way to distill the knowledge contained in one …

Attributed Social Network Embedding, Lizi Liao, Xiangnan He, Hanwang Zhang, Tat-Seng Chua

Research Collection School Of Computing and Information Systems

Embedding network data into a low-dimensional vector space has shown promising performance for many real-world applications, such as node classification and entity retrieval. However, most existing methods focused only on leveraging network structure. For social networks, besides the network structure, there also exists rich information about social actors, such as user profiles of friendship networks and textual content of citation networks. These rich attribute information of social actors reveal the homophily effect, exerting huge impacts on the formation of social networks. In this paper, we explore the rich evidence source of attributes in social networks to improve network embedding. We …

Object Localization, Segmentation, And Classification In 3d Images, Allan Zelener

Dissertations, Theses, and Capstone Projects

We address the problem of identifying objects of interest in 3D images as a set of related tasks involving localization of objects within a scene, segmentation of observed object instances from other scene elements, classifying detected objects into semantic categories, and estimating the 3D pose of detected objects within the scene. The increasing availability of 3D sensors motivates us to leverage large amounts of 3D data to train machine learning models to address these tasks in 3D images. Leveraging recent advances in deep learning has allowed us to develop models capable of addressing these tasks and optimizing these tasks jointly …

Recurrent Neural Networks And Their Applications To Rna Secondary Structure Inference, Devin Willmott

Theses and Dissertations--Mathematics

Recurrent neural networks (RNNs) are state of the art sequential machine learning tools, but have difficulty learning sequences with long-range dependencies due to the exponential growth or decay of gradients backpropagated through the RNN. Some methods overcome this problem by modifying the standard RNN architecure to force the recurrent weight matrix W to remain orthogonal throughout training. The first half of this thesis presents a novel orthogonal RNN architecture that enforces orthogonality of W by parametrizing with a skew-symmetric matrix via the Cayley transform. We present rules for backpropagation through the Cayley transform, show how to deal with the Cayley …

Multilabel Learning For The Online Transient Stability Assessment Of Electric Power Systems, Peyman Beyranvand, Veysel Murat İstemi̇han Genç, Zehra Çataltepe

Turkish Journal of Electrical Engineering and Computer Sciences

Dynamic security assessment of a large power system operating over a wide range of conditions requires an intensive computation for evaluating the system's transient stability against a large number of contingencies. In this study, we investigate the application of multilabel learning for improving training and prediction time, along with the prediction accuracy, of neural networks for online transient stability assessment of power systems. We introduce a new multilabel learning method, which uses a contingency clustering step to learn similar contingencies together in the same multilabel multilayer perceptron. Experimental results on two different power systems demonstrate improved accuracy, as well as …

Modeling And Mapping Location-Dependent Human Appearance, Zachary Bessinger

Theses and Dissertations--Computer Science

Human appearance is highly variable and depends on individual preferences, such as fashion, facial expression, and makeup. These preferences depend on many factors including a person's sense of style, what they are doing, and the weather. These factors, in turn, are dependent upon geographic location and time. In our work, we build computational models to learn the relationship between human appearance, geographic location, and time. The primary contributions are a framework for collecting and processing geotagged imagery of people, a large dataset collected by our framework, and several generative and discriminative models that use our dataset to learn the relationship …

Pseudorehearsal In Actor-Critic Agents With Neural Network Function Approximation, Vladimir Marochko, Leonard Johard, Manuel Mazzara, Luca Longo

Articles

Catastrophic forgetting has a significant negative impact in reinforcement learning. The purpose of this study is to investigate how pseudorehearsal can change performance of an actor-critic agent with neural-network function approximation. We tested agent in a pole balancing task and compared different pseudorehearsal approaches. We have found that pseudorehearsal can assist learning and decrease forgetting.

Hybrid Privacy-Preserving Clinical Decision Support System In Fog-Cloud Computing, Ximeng Liu, Robert H. Deng, Yang Yang, Ngoc Hieu Tran, Shangping Zhong

Research Collection School Of Computing and Information Systems

In this paper, we propose a framework for hybrid privacy-preserving clinical decision support system in fog cloud computing, called HPCS. In HPCS, a fog server uses a lightweight data mining method to securely monitor patients' health condition in real-time. The newly detected abnormal symptoms can be further sent to the cloud server for high-accuracy prediction in a privacy-preserving way. Specifically, for the fog servers, we design a new secure outsourced inner-product protocol for achieving secure lightweight single-layer neural network. Also, a privacy-preserving piecewise polynomial calculation protocol allows cloud server to securely perform any activation functions in multiple-layer neural network. Moreover, …

Estimating Meteorological Visibility Range Under Foggy Weather Conditions: A Deep Learning Approach, Hazar Chaabani, Naoufel Werghi, Faouzi Kamoun, Bilal Taha, Fatma Outay, Ansar Ul Haque Yasar

All Works

© 2018 The Authors. Published by Elsevier Ltd. Systems capable of estimating visibility distances under foggy weather conditions are extremely useful for next-generation cooperative situational awareness and collision avoidance systems. In this paper, we present a brief review of noticeable approaches for determining visibility distance under foggy weather conditions. We then propose a novel approach based on the combination of a deep learning method for feature extraction and an SVM classifier. We present a quantitative evaluation of the proposed solution and show that our approach provides better performance results compared to an earlier approach that was based on the combination …

Symbolic Interpretation Of Artificial Neural Networks Using Genetic Algorithms, Dounia Yedjour, Abdelkader Benyettou, Hayat Yedjour

Turkish Journal of Electrical Engineering and Computer Sciences

The knowledge acquired during the learning of artificial neural networks (ANNs) is coded as values in synaptic weights, which makes their interpretations difficult, hence the name of the black box. The aim of this work is to provide a comprehensible interpretation of the ANN's decisions by extracting symbolic rules. We improve the performance of our extraction algorithm by combining the ANN with a genetic algorithm. Misleading rules whose support and confidence values are less than fixed thresholds are removed and, as a result, the comprehensibility is improved. The extracted rules are evaluated and compared with other works. The results show …

Visual Odometry Using Convolutional Neural Networks, Alec Graves, Steffen Lim, Thomas Fagan, Kevin Mcfall Phd.

The Kennesaw Journal of Undergraduate Research

Visual odometry is the process of tracking an agent's motion over time using a visual sensor. The visual odometry problem has only been recently solved using traditional, non-machine learning techniques. Despite the success of neural networks at many related problems such as object recognition, feature detection, and optical flow, visual odometry still has not been solved with a deep learning technique. This paper attempts to implement several Convolutional Neural Networks to solve the visual odometry problem and compare slight variations in data preprocessing. The work presented is a step toward reaching a legitimate neural network solution.

Online Learning With Nonlinear Models, Doyen Sahoo

Dissertations and Theses Collection (Open Access)

Recent years have witnessed the success of two broad categories of machine learning algorithms: (i) Online Learning; and (ii) Learning with nonlinear models. Typical machine learning algorithms assume that the entire data is available prior to the training task. This is often not the case in the real world, where data often arrives sequentially in a stream, or is too large to be stored in memory. To address these challenges, Online Learning techniques evolved as a promising solution to having highly scalable and efficient learning methodologies which could learn from data arriving sequentially. Next, as the real world data exhibited …

Deepfacade: A Deep Learning Approach To Facade Parsing, Hantang Liu, Jialiang Zhang, Jianke Zhu, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

The parsing of building facades is a key component to the problem of 3D street scenes reconstruction, which is long desired in computer vision. In this paper, we propose a deep learning based method for segmenting a facade into semantic categories. Man-made structures often present the characteristic of symmetry. Based on this observation, we propose a symmetric regularizer for training the neural network. Our proposed method can make use of both the power of deep neural networks and the structure of man-made architectures. We also propose a method to refine the segmentation results using bounding boxes generated by the Region …

Deep Learning On Lie Groups For Skeleton-Based Action Recognition, Zhiwu Huang, C. Wan, T. Probst, Gool L. Van

Research Collection School Of Computing and Information Systems

In recent years, skeleton-based action recognition has become a popular 3D classification problem. State-of-the-art methods typically first represent each motion sequence as a high-dimensional trajectory on a Lie group with an additional dynamic time warping, and then shallowly learn favorable Lie group features. In this paper we incorporate the Lie group structure into a deep network architecture to learn more appropriate Lie group features for 3D action recognition. Within the network structure, we design rotation mapping layers to transform the input Lie group features into desirable ones, which are aligned better in the temporal domain. To reduce the high feature …

Cascaded Facial Detection Algorithms To Improve Recognition, Edmund Yee

Master's Projects

The desire to be able to use computer programs to recognize certain biometric qualities of people have been desired by several different types of organizations. One of these qualities worked on and has achieved moderate success is facial detection and recognition. Being able to use computers to determine where and who a face is has generated several different algorithms to solve this problem with different benefits and drawbacks. At the backbone of each algorithm is the desire for it to be quick and accurate. By cascading face detection algorithms, accuracy can be improved but runtime will subsequently be increased. Neural …

Hand Gesture Recognition For Sign Language Transcription, Iker Vazquez Lopez

Boise State University Theses and Dissertations

Sign Language is a language which allows mute people to communicate with other mute or non-mute people. The benefits provided by this language, however, disappear when one of the members of a group does not know Sign Language and a conversation starts using that language. In this document, I present a system that takes advantage of Convolutional Neural Networks to recognize hand letter and number gestures from American Sign Language based on depth images captured by the Kinect camera. In addition, as a byproduct of these research efforts, I collected a new dataset of depth images of American Sign Language …

Impact Of Reviewer Social Interaction On Online Consumer Review Fraud Detection, Kunal Goswami, Younghee Park, Chungsik Song

Faculty Publications

Background Online consumer reviews have become a baseline for new consumers to try out a business or a new product. The reviews provide a quick look into the application and experience of the business/product and market it to new customers. However, some businesses or reviewers use these reviews to spread fake information about the business/product. The fake information can be used to promote a relatively average product/business or can be used to malign their competition. This activity is known as reviewer fraud or opinion spam. The paper proposes a feature set, capturing the user social interaction behavior to identify fraud. …

Ai Education: Deep Neural Network Learning Resources, Todd W. Neller

Computer Science Faculty Publications

In this column, we focus on resources for learning and teaching deep neural network learning. Many exciting advances have been made in this area of late, and so many resources have become available online that the flood of relevant concepts and techniques can be overwhelming. Here, we hope to provide a sampling of high-quality resources to guide the newcomer into this booming field. [excerpt]

An Alternative Approach To Training Sequence-To-Sequence Model For Machine Translation, Vivek Sah

Honors Theses

Machine translation is a widely researched topic in the field of Natural Language Processing and most recently, neural network models have been shown to be very effective at this task. The model, called sequence-to-sequence model, learns to map an input sequence in one language to a vector of fixed dimensionality and then map that vector to an output sequence in another language without any human intervention provided that there is enough training data. Focusing on English-French translation, in this paper, I present a way to simplify the learning process by replacing English input sentences by word-by-word translation of those sentences. …

Towards A Relative-Pitch Neural Network System For Chorale Composition And Harmonization, Samuel P. Goree

Honors Papers

Computational creativity researchers interested in applying machine learning to computer composition often use the music of J.S. Bach to train their systems. Working with Bach, though, requires grappling with the conventions of tonal music, which can be difficult for computer systems to learn. In this paper, we propose and implement an alternate approach to composition and harmonization of chorales based on pitch-relative note encodings to avoid tonality altogether. We then evaluate our approach using a survey and expert analysis, and find that pitch-relative encodings do not significantly affect human-comparability, likability or creativity. However, an extension of this model that better …