Python toolkit
DYANA
DYadic Annotation of Naturalistic Audio: a transparent pipeline for turning messy two-person recordings into inspectable conversational annotations, evidence traces, and TextGrid-ready outputs.
Software
Tools for speech, annotation, and neural data.
Current software projects across dyadic audio analysis, speech annotation, voice and language feature extraction, mixed-effects EEG workflows, and neural time-series modeling. Each row keeps the core tool on the left and practical context on the right.
Desktop application
Movak
A modern speech-annotation workspace for the stage after automatic alignment, where researchers need to inspect signals, correct boundaries, relabel tiers, and curate trustworthy annotations.
Python toolkit
VoxAtlas
A modular conversational feature-extraction toolkit that turns messy speech data into organized acoustic, linguistic, and alignment-based layers through reusable extractors and dependency-aware pipelines.
Python package
LmeEEG
A lightweight Python workflow for mixed-effects mass-univariate EEG analysis, built to model trial-wise predictors, marginalize random-intercept structure, and run corrected spatiotemporal inference transparently.
Scientific library
NeuroWeave
A Python library for explicit temporal encoding models, where feature spaces, lag operators, estimators, models, and validation are kept separate and inspectable.