Kun Woo Cho | 조건우

In-ear Sensing System for Smartphone User Authentication

The Embedded Sensing and Computing Lab, University at Buffalo (08/2017 - present)

• Lead Researcher
• Implementation of an authentication system based on the reflection of acoustic waves in the earphone.

Smartphone based Adaptive Platform to Track User Emotions

The Networks and Operating Systems Lab, University of Cambridge (06/2017 - 08/2017)

• Lead Researcher
• EmotionSense is a mobile sensing platform that was developed in 2010 by Kiran et al. It focused on measuring social group dynamics using a variety of sensors. In comparison to the previous model, I redeveloped EmotionSense app with an exclusive focus on the sound recognition via a microphone sensor in the smartphones. My proposed system comprises four modules: sound signal acquisition, signal pattern analysis, efficient power management, and interactive user interfaces. In summary, the project goals include the following: (1) how to discriminate the silent signal and non-silent signal; (2) how to differentiate the speech and ambient sound for non-silent signals; (3) how to recognize the specific-emotion in the speech sound and further categorize this specific-emotion into the broad emotion, such as happiness, sadness, fear, anger, or neutral; (4) how to achieve an energy-efficient system via adaptive duty cycle algorithm and optimization techniques; (5) how to design the user interface in the way that user can easily understand the trend of emotions over time.

User Authentication for Smart Headwear through Cancelable Brain Biometrics

The Embedded Sensing and Computing Lab, NSF REU Program (05/2016 - 08/2017)

• Lead Researcher
• With a growing popularity of virtual reality, head-mounted displays (HMD) have been widely used in many applications, such as civilian-commercials, military, and medicine. However, a practical authentication system in smart headwear devices is still challenging and struggles from an appropriate balance between security and usability. We explored a secure, usable, and cancelable user authentication method using brain-computer interface. Specifically, we present a new psycho-physiological protocol for cancelable and trustworthy user authentication of HMD through an event-related potential (ERP). In this study, we attempt to answer the following research challenge in an experimental and theoretical combined manner: (1) how to generate effective ERP biosignals; (2) how to acquire and analyze unique features in ERP biosignals for authentication; (3) how to reset and change brain biometrics when the current biometric credential is divulged. To the best of our knowledge, this project is the first in-depth research study on truly cancelable brain biometrics.

EEG-based Personal Identification Using Mallows Distance

The Embedded Sensing and Computing Lab, University at Buffalo (09/2015 - 11/2015)

Early Screening Approach to Autism Spectrum Disorder Using Discrete Gaze Pattern

The Embedded Sensing and Computing Lab, University at Buffalo (11/2015 - 05/2016)

• Lead Researcher
• Early detection of children with autism spectrum disorder (ASD) is essential in maximizing the positive impact of autism intervention. However, the diagnosis of ASD are difficult to receive in a timely manner due to the needs for highly specialized settings and trained professionals. In contrast to current methods, we introduce a new paradigm for the diagnosis and assessment of ASD: a novel ASD screening method, namely Gaze-Wasserstein, that is non-invasive, fast, and widely accessible. Based on the gaze tracking and analysis, Gaze-Wasserstein provides objective gaze pattern-based measurements for ASD screening, which can be deployed on any mobile device with an embedded camera. In Gaze-Wasserstein, a modified 1st Wasserstein distance is implemented for global gaze pattern matching, and social stimuli are deliberately designed to elicit abnormal gaze behaviors of the children with ASD. Our evaluation results from the study substantiate the feasibility and time-efficiency of our proposed method. Moreover, our exploratory study investigates the effects of facial expression and image resolutions on the gaze fixation consistency of children with ASD.

Early Childhood Education Through Cyber-learning

The Embedded Sensing and Computing Lab and the Department of Learning and Instruction, University at Buffalo (07/2016 - 09/2016)

• Research Assistant
• We investigated the correlation between children's touch motions on the interactive eBook and their inference & critical thinking skills. As a reserach assistant, I developed a program that simulates the page flipping motion and touch motion on the hotspots, including main characters and interactive objects. This program also counts the touch frequency for each hotspot.

Thermal Handprint-based Personal Identification Using Heat-Earth Mover's Distance

The Embedded Sensing and Computing Lab, SUNY Passport Program (07/2015 - 09/2015)

• Lead Researcher
• Recently, handprint-based recognition system has been widely applied for security and surveillance purposes. The success of this technology has demonstrated that handprint is a good approach to perform forensic identification. However, existing identification systems are nearly based on the handprints that could be easily prevented. In contrast to earlier works, we exploit the thermal handprint and introduce a novel distance metric for thermal handprint dissimilarity measure, called Heat-Earth Mover’s Distance (HEMD). The HEMD is designed to classify heat-based handprints that can be obtained even when the subject wears a glove. HEMD can effectively recognize the subjects by computing the distance between point distributions of target and training handprints. Through a comprehensive study, our identification system demonstrates the performance even with the handprints obtained by the subject wearing a glove. Further, we investigated the effects of surface emissivity on the duration of thermal handprints.