• Efficient View-Based 3D Reflection Symmetry Detection

  • Symmetry is an important clue for geometry perception: it is not only in many man-made models, but also widely exists in the nature. Symmetry has been used in many applications such as: 3D alignment, shape matching, remeshing, 3D model segmentation and retrieval. The objective of this project is to propose and develop an efficient method to automatically detect 3D symmetry by adopting a view-based approach.

  • Kinect-Assisted 3D Sketch-based 3D Model Retrieval

  • Existing sketch-based 3D model retrieval systems rely on 2D sketching technology. A 2D sketch lacks the 3D shape information of the object it represents, which generates a large semantic gap between the iconic representation of the 2D sketch and the accurate representation of the 3D model behind. It is this gap that makes 2D sketch-based 3D modeling retrieval a very challenging task. The goal of this project is to use 3D sketches for 3D model retrieval and develop a 3D sketch-based 3D model retrieval system in order to bridge the above semantic gap.

  • Semantic Tree-based 3D Model Retrieval Using 2D Sketch Queries

  • Effectively and efficiently retrieving relevant 3D models for a 2D sketch query is important for various related applications. Due to the big semantic gap existing between rough sketch representation and accurate 3D model coordinates, sketch-based 3D model retrieval is one of the most challenging research topics in the field of 3D model retrieval. To bridge the semantic gap, the objective of this project is to perform semantics-driven sketch-based 3D model retrieval based on a semantic tree.

  • Deep Learning for Sketch Recognition

  • Sketching is the only way for most people to render visual content. However, due to the special characters of the sketch, most traditional feature extraction techniques, like SIFT, have poor performance on these tasks. The objective of this project is to apply deep learning approach on sketch understanding and sketch scene recognition.

    Sensor Network

  • Wearable Sensor Network

  • With the coming of Google glasses, smartwatches, wristbands, and smartphone, etc., the wearables have changed and will change our lives tremendously by connecting people anywhere anytime in a scale that has never seen before. They enable us to obtain real-time data about people and environment so that useful applications can be developed to assist people. The success of the applications depends on the accuracy of data. However, due to different reasons, data gathered may be subject to noise, missing information, and anomaly. The objective of this project is to conduct experiments with several wearable sensors, analyze the gathered data and custom-design new data processing methods addressing the above challenges using data mining and machine learning techniques.

  • Detecting Driver Drowsiness using Wireless Wearables

  • The National Highway Traffic Safety Administration data show that drowsy driving causes more than 100,000 crashes a year. In order to prevent these devastating accidents, it is necessary to build a reliable driver drowsiness detection system which could alert the driver before a mishap happens. In the literature, the drowsiness of a driver can be measured by vehicle-based, behavior-based, and physiology-based approaches. Comparing with the vehicle-based and behavior-based measurements, the physiological measurement of drowsiness is more accurate. With the latest release of wireless wearable devices such as biosensors that can measure people's physiological data, the objective of this project to explore the possibility of designing a user-friendly and accurate driver drowsiness detection system using wireless wearables.

    Data Mining

  • Parallel Learning to Rank

  • The objective of this project is to leverage parallelization to improve the efficiency of learning to rank techniques in the big data setting.

  • Iterative Intercolumnar Correlation and Its Applications to Clustering

  • The objective of this project is to investigate a mathematical conjecture about iterative intercolumnar correlation and explore its applications in clustering.

    High Performance Computing

  • Autotuning Tools for HPC Systems

  • Autotuning has emerged as a promising strategy for harnessing the computational capabilities of future Exascale systems. While autotuning has been quite successful for specific domains, transcending the model from domain-specific kernels to general applications has been difficult. The autotuning effort at Texas State aims to tackle the search space problem by focusing on the feedback aspects. Efficiency in our framework is achieved through enhanced knowledge of the problem domain, program features and architectural characteristics. To this end, we are developing tools and models that allow specification, collection and synthesis of extremely fine-grain data related to performance and power consumption. We are also developing novel search heuristics and supervised learning models that take advantage of this rich information. Our goal is to provide an open-source tool-chain for automatic performance tuning that integrates each of the component tools to deliver portable performance for broad range applications.

    Internet of Thing (IOT) and Web Service

  • Social-PPM: Personal Processes Sharing and Recommendation

  • The rise in popularity of various social network applications has brought the opportunities for Internet users to share and reuse a plethora of things like images, videos, datasets, ideas, interests, reviews etc. However, currently there is no effective way to share personal experiences such as the process of filing a personal income tax return or the process of applying a visa. The objective of this project is to provide a social-aware process model and its implementation as a social network application that empowers users to create, to execute, and to share personal experiences within a social network at anytime and at anywhere.

  • A Flexible Internet of Things Middleware

  • The Internet of Things (IoT) is a rapidly growing system of physical, virtual, and social sensors, enabling an advanced information gathering, interpretation and monitoring. However, IoT must be supported by a middleware that allows IoT consumers and IoT application developers to interact in a user-friendly way, despite the differences in each user's perspective of IoT system. The objective of this project is to bridge the gap between IoT consumers and IoT application developers in consuming IoT data and composing IoT applications.

  • Sensing, Collecting and Predicting Drunkenness using Smartwatch and Smartphone

  • Internet of Things (IoT) is a domain that represents the next most exciting technological revolution since Internet. IoT will bring endless opportunities and impact every corner of our planet. In the healthcare domain, IoT promises to bring personalized health tracking and monitoring ever closer to the consumers. Modern smartphones and related devices now contain more sensors than ever before. Data from sensors can be collected ever more easily and more accurately. This project aims to combine machine learning techniques and innovative data collection system for smartwatches to predict Blood Alcohol Content (BAC) non-invasively and in real time.

  • ServiceXplorer: A Similarity-based Web Service Search Engine

  • Most existing Web service search engines employ keyword search over databases, which computes the distance between the query and the Web services over a fixed set of features. Such an approach often results in incompleteness of search results. The Earth Mover's Distance (EMD) has been successfully used in multimedia databases due to its ability to capture the differences between two distributions. However, computing EMD is computationally expensive. In this project, we aim to improve the existing keyword-based search techniques for Web services. In particular, we explore using EMD for many-to-many partial match between the contents of the query and the service attributes.

  • Systematic Collection of Computer Science PhD Job Postings

  • This project aims to develop a computer program to automatically collect and analyze job-posting data from various high-tech companies and technology job-search websites for establishing the job market for Computer Science Ph.D. holders. The collected data must be automatically analyzed for duplicates and then queried for detailed analysis in terms of most needed skill set or location based demand.