fMRI · Naturalistic Speech · Computational Models · Auditory Neuroscience
Examining neural mechanisms behind cognitive processes, with a focus on naturalistic stimuli related to hearing and speech.
Utilizing computational methods to model auditory and speech processing in the brain — bridging signal processing and neuroscience.
Employing fMRI techniques to investigate the neural basis of auditory and speech functions, with clinical applications in mind.
Studying how auditory cortex subdivisions and the Multiple Demand Network reconfigure dynamically during challenging listening conditions.
Software built to move research forward — and studies that use it to map how the brain hears.
Using fMRI with naturalistic speech stories at varying SNRs to map how auditory cortex and the Multiple Demand Network reconfigure under degraded listening conditions.
The right insula emerges as a key hub — bridging auditory and cognitive systems when speech becomes hard to follow.
A Snakemake/Snakebids pipeline for preprocessing and analyzing epilepsy fMRI data. Runs end-to-end from raw BIDS input through fMRIprep, FreeSurfer surface extraction, fsLR 32k projection, and inter-subject correlation (ISC) analysis across four parcellation atlases.
Scales to 100 concurrent SLURM jobs on the Alliance Canada cluster. Supports Desikan, Glasser, Schaefer, and Yan atlases.
An audio signal processing toolkit for hearing and communication research. Triton organizes stimulus preparation around a project-centric workflow — import and auto-normalize audio, apply multi-label categorization, build reproducible processing pipelines with parameter sweeps, and generate or degrade stimuli with precise control over level, noise, and envelope shape. Every step is logged with JSON sidecar provenance.
Interfaces: Streamlit GUI · CLI · Python API. Audio ops: SNR mixing, 4 ramp envelopes (linear/exp/log/cosine), RMS normalization. Visualization: STFT, Mel, and CQT spectrograms. Stack: Python + FFmpeg + librosa, env-locked with Pixi. MIT licensed.
Hobby builds, photography, and whatever else keeps the hands busy outside research hours.
Comprehensive keyboard shortcut reference for tmux — sessions, windows, panes, copy mode, and power tips. Printable A4 layout.
View →A daily journaling app that transforms your writing into verse styled after Ferdowsi's Shahnameh — built with Google AI Studio.
View →A free, browser-based teleprompter — supports .txt, .md, and .docx, mirror mode, remote clicker mapping, and adjustable auto-scroll speed.
View →Photos will go here.