Submitted to EICT 2025

Submitted to TPS 2025

Submitted to ICMLA 2025

Published to KSU symposium 2025

Published to IEEE VR 2025

Masters Thesis, Published to KSU in 2024

Masters thesis, published to IEEEVR 2024, funded by US Department of Defense

Masters course research, published to KSU in 2023

AIAST-lab, Kennesaw State University

Duties/works

• Research: Machine Learning privacy, developing robust model against adversarial attack

HIX-lab, Kennesaw State University

Duties/works

• Research: Cybersickness prediction with health data (HR, EDA, Head and Eye movement)
• Developed model using RNN, GRU, LSTM, CNN-LSTM, Attention-LSTM, Transformer, DeepTCN
• Deployed DNN models to Snapdragon chipset with UNITY SDK in Meta Quest 3
• Worked with Hp Omnicept Reverb G2, VR Headsets, Emotibit, ESP32, Virtual-Reality, Mixed-Reality
• Developed framework for real-time Cybersickness detection and mitigation using CS time series data
• Participated to several user study as a participant including labs and Facebook research team

HIX-lab, Kennesaw State University

Duties/works

• Research: Cybersickness prediction with health data (HR, EDA, Head and Eye movement)
• Developed model using RNN, GRU, LSTM, CNN-LSTM, Attention-LSTM, Transformer, DeepTCN
• Deployed DNN models to Snapdragon chipset with UNITY SDK in Meta Quest 3
• Worked with Hp Omnicept Reverb G2, VR Headsets, Emotibit, ESP32, Virtual-Reality, Mixed-Reality
• Developed framework for real-time Cybersickness detection and mitigation using CS time series data
• Participated to several user study as a participant including labs and Facebook research team

Chowagiken

Duties/works

• Developed Car Image Segmentation model using Yolov6- In Web application deployed that model, used Flask for API. For production, used Docker to dockerize the whole application, deployed on the client’s GCP computing server and also used GitHub Action for CI/CD.
• Developed Image Anomaly detection application- In this project, classified defected potatoes using Pathcore and Padim, for auto segmenting used DINO and SAM. Made Industry graded documentation for client
• Fined tuned Yolov8, YoloX, RTMDet on several image dataset (weed dataset, mosquito feeding, fire detection)
• Developed functionalities for SMPL model 3d & depth map generation in Blender
• Learned punctuality, maintaining delivery deadline, and politeness in Japanese working culture

Undergraduate Thesis and Project

Contribution: In this thesis water data was collected with sensors and sent to server and retrieved with android application. Water flow, volume, used volume, pH, turbity and flushing process was done.