DIMS documentation

DIMS puts every signal from a recorded session on one timeline, beside the video and transcript they came from, with the analyses that say how two signals relate. Click a peak in a time series and the video jumps to the moment that produced it.

It is for researchers who already have time-aligned measures — motion tracking, physiology, gaze codes, annotation tiers — and want to get from a statistical pattern back to the raw record that produced it.

What DIMS is not: it is not a preprocessing pipeline. It does not clean signals, estimate synchrony, or align clocks. It displays what you give it, which means it displays a misaligned signal faithfully as a misaligned signal.

The video is optional. The core is driven by timestamps, not by a recording — a study with no .mp4 anywhere still gets the time series, RQA, cross-RQA, cross-wavelet and ELAN tabs, all on the same timeline. That covers foot-pressure and other insole sensing, audio-only corpora, motion capture, EMG, eye-tracking, and any study whose video was stripped for de-identification. What you lose is the video panel and the ability to click from a peak to the moment on screen; nothing else changes.

See a live dashboard → Click the timeline; the video and every chart follow.

This site is for two kinds of reader: researchers who want to explore their own recordings with DIMS, and contributors who want to add a tab, an analysis, or a fix. The nav above splits along that line — Welcome and For no-code dashboard get you to your first dashboard without writing code; Overview of existing analysis modules shows what DIMS already computes; Extending DIMS is the contracts and reference material for changing the code itself.

---

Published at https://dims-network.github.io/dims/, built from this directory once per release — a page on the site and the file beside the code are the same text, at the version you are reading. See Citing DIMS for the paper and the code.