Student Poster Competition
1st Place Winner
Stevens Institute of Technology
Chen Wang received his BS and MS degrees from the University of Electronic Science and Technology of China, in 2009 and 2012. He is currently pursuing a PhD degree in the Electrical and Computer Engineering Department at Stevens Institute of Technology under the supervision of Professor Yingying Chen. His research interests include mobile computing, cyber security and privacy, and smart healthcare. He received the Best Paper Award from the ACM Conference on Information, Computer and Communications Security (ASIACCS) 2016 and the Best Paper Award from the IEEE Conference on Communications and Network Security (CNS) 2014.
Abstract Title: Fine-grained Sleep Monitoring Using Smartphones
Authors: Yanzhi Ren, Chen Wang, Yingying Chen and JieYang
Sleep monitoring has drawn increasing attention as the quality and quantity of sleep are important for maintaining a person’s health and well-being. For example, inadequate and irregular sleep is usually associated with serious health problems such as fatigue, depression and cardiovascular disease. Traditional sleep monitoring systems, such as PSG, involve wearable sensors with professional installations, and thus are limited to clinical usage. Recent work in using smartphone sensors for sleep monitoring can detect several events related to sleep, such as body movement, cough and snore. Such coarse-grained sleep monitoring however is unable to detect the breathing rate which is a vital sign and health indicator. This work presents a fine-grained sleep monitoring system which is capable of detecting the breathing rate by leveraging smartphones. Our system exploits the readily available smartphone earphone that is placed close to the user to capture the breath sound reliably. Given the captured acoustic signal, our system performs noise reduction to remove environmental noise and then identifies the breathing rate based on the signal envelope detection. Our experimental evaluation of six subjects over a six-month time period demonstrates that the breathing rate monitoring is highly accurate and robust under various environments. This strongly indicates the feasibility of using the smartphone and its earphone to perform continuous and noninvasive fine-grained sleep monitoring.
2nd Place Winner
Electrical and Computer Engineering
Rutgers University, Class of 2017
Eric Wengrowski is a Computer Vision PhD student at Rutgers University in the Electrical & Computer Engineering Department. His research adviser is Professor Kristin J. Dana. Eric’s research interests include computational photography, imaging, and machine learning. Previously, Eric held summer internships at Kitware, Microsoft Research, a San Francisco Tech Startup, AT&T Labs, and the US Army Corps of Engineers.
Abstract Title: Reflection Matching for Detecting Photograph Manipulations
Modern photo editing software like Photoshop has made it difficult to trust the authenticity of digital images. In response, image forensic techniques have emerged to detect photo manipulations. When jpeg-based authentication methods fail, a photograph’s integrity can be examined by checking if scene geometry is preserved in reflections. Environmental reflections contain useful information about the geometry and photometry of objects in a scene. Ersatz reflective geometry might fool the human eye, but image integrity can be verified or questioned if reflective correspondence is known. Traditionally, human input was needed to manually annotate reflective correspondences, a tedious and error-prone process. We introduce a new algorithm for automatically finding correspondences between scene objects and their reflections, with manual specification of only reflection regions instead of individual points. It is assumed that images contain scene elements that are imaged both directly and indirectly through planar reflection. Results are presented on interesting cases, both successes and failures where automated correspondence is very difficult. We also discuss the motivation and challenges associated with reflection correspondence within single images.
3rd Place Winner
Computer Science and Psychology (Double Major)
Fairleigh Dickinson University, Class of 2018
Kathleen Falcon is a junior at Fairleigh Dickinson University in Madison, NJ. She is double majoring in Psychology & Computer Science with a concentration in Cybersecurity and a minor in Mathematics. She is interested in furthering her education in Computer Science with graduate school in her pursuit of a career in Cybersecurity. Her areas of research interest include Arduino and Raspberry Pi programming, Blockchain, and the Internet of Things. Her hobbyist interests include quadcopter/drone technology, 3D printing, and building desktop computers.
Abstract Title: IoT: Mobile Device Network Security using Prototyping Platforms
Internet of Things (IoT), a network of physical objects embedded with electronics, software, sensors, and network connectivity; once seemed like science fiction, is now paving the way for a modern, smart, connected home/world to become a reality. Today, there are more connected devices than people and Cisco predicts that there will be over 50 billion devices by 2020. Many appliances and gadgets in use today are already “smart” (network/Internet- accessible) as IoT has expanded into healthcare (smart monitoring, diagnoses and treatment), automotive (smart cars), control & automation (smart industries), real estate (smart homes), public utilities (smart grids, water/waste management, transportation, etc.). Machines talking to machines (without human intervention) - a key feature of IoT, has the potential to save millions of dollars and benefit society. However, such widespread networking opens up daunting security challenges that can modify medical devices, jeopardize automobiles, shut down the electric grid, cause traffic mayhems, and destroy homes. Apart from network traffic, IoT devices generate vast amounts of data demanding data security improvements. Also as they proliferate, we will need to ensure that data security controls and practices scale accordingly. Using prototyping platforms (such as the Arduino) and interconnected devices, we developed a “smarthome” control application with context-aware security using situational information and mutual authentication (between user & device and between devices) to provide safe infrastructure for user and device to work seamlessly. In the future we wish to look into Blockchain technology (recently used by financial firms to simplify trade finance processes) for IoT Cybersecurity.
3rd Place Winner
Fairleigh Dickinson University
Mital began her research this fall with advisor Dr. Erdal Kose in the Computer Science department of Fairleigh Dickinson University on learning and applying artificial intelligence techniques. Specifically her ideas are the implementation of Deep Learning Neural Network on Spark framework for driverless cars.
Abstract Title: Use of big data analytics and artificial intelligence for autonomous decisions
Big data is getting successful in gathering and storing massive conglomerated data by use of Hadoop technologies. But, day by day the data is getting complex and becoming available from so many sources. Embedding the fields of big data, machine learning, artificial intelligence, expert systems and analytics will lead to use of intelligent brain for making autonomous, optimal, productive and faster decisions for humans. AI adds an intelligence layer to big data to tackle complex analytical tasks much faster than humans can cope up. Big data provides a huge amount of data for AI to continue to learn and evolve. Chat bots and virtual personal assistants, Google’s self-driving car, recommendation systems are some of the examples of AI for self-guided systems. In this poster we shall discuss the different applications and technologies of AI and big data currently being used by the different companies like EBay, Amazon, Netflix to increase their customer service and marketing.
3rd Place Winner
Kevin Scott Miller
MS Computer Science
Montclair State University, Class of 2017
Kevin S. Miller is an MS Computer Science student at Montclair State University, where he also received his BS CS in 2015. He earned an AS degrees in Electronics Technology and Computer Technology in 1983 and in Computer Science in 2014. After working as a technician and self-taught programmer for 30 years in industries ranging from Astronomy to Private Banking, Kevin decided to complete his formal education in 2012.
Abstract Title: SDR Toolkit for IoT Development
There is a need for high-efficiency short-range wireless communications to connect IoT devices that have low to medium security requirements. A hardware/software tool was developed to help IoT product developers quickly and easily develop radio frequency (RF) communication systems for IoT devices where previously this was a manual, one-off process. The tool uses Software Defined Radio (SDR) and focuses on On-Off-Keying (OOK) modulation. It can be used by persons with limited knowledge of RF to analyze existing devices and capture its characteristics, which can be used to create and transmit new messages, in effect spoofing it. New device definitions can be implemented in low-cost, off-the-shelf hardware for production. OOK has been found to be very efficient at binary RF communications because the transmitter is only powered when a “1” is being transmitted. This efficiency translates into a battery life of up to one year. Implementations of this system could include arrays of sensors that periodically transmit data to a traditionally-powered Internet-connected receiver. Another possible use of this system could be low-cost small transmitters to track animal movements in a defined area. Receivers placed around the area could record the time and signal strength of the transmissions. Software would be used to analyze the data and plot the animal’s movements. Because the RF transmissions have a specific range, the opportunity to intercept, modify or spoof communications would be limited. For sensitive data, rolling codes and/or public/private key encryption could be used for encoding before modulating with OOK.