Projects

Using RIFT to measure spatial attention allocation via EEG hemisphere asymmetry. Clinical target: diagnosing and monitoring Unilateral Spatial Neglect in stroke patients.

Designing adaptive signal architecture for Gabb's K-12 language learning platform in XR environments. Research focus: struggle detection, spatial memory, attention and cognitive load detection, neuroadaptive interaction.

Designed and conducted user studies in Unity-3D; developed DL-based classifiers for neural intention detection to improve gaze-based interactions. Accepted to the ACM CHI 2026 main conference.

Developed GAN-generated images for human-in-the-loop experiment, integrated physiological data (ERP and Gaze) with perceptual and attentional metrics, precise stimulus timing, and ERP-compatible event logging.

Designed multimodal BCI for paralyzed patients to control IoT devices, robotic hand prototype via integrating real-time signals, EEG and eye-tracking, with AI-enhanced health monitoring. Won first place at TEKNOFEST 2025.

Evaluated local LLM performance via comparative RAG and hardware-aware pipelines, integrating the security-tuned Llama-3.1 reasoning model with MinerU for computer vision-based multimodal context extraction.

Developed full-stack application portal with Flask REST API, React frontend, PostgreSQL database, and LLM-integrated matching/chatbot systems.

Showed that applying multivariate amplitude-binned envelope to regression based TRF modeling increases neural response prediction. Currently extending the work with YAMnet sound classification to detect causal effects between speech and non-speech sounds.

Modeling caffeine effects of gap-junction conductance on electrically coupled neurons. Found causal effects on alertness and spike features at the single-cell level.