Data tables and statistics

3D Data Analyzer ↗

Visualises, analyses and edits 3D data: Qualisys motion capture and body, hand, face and point tracking from video. Trajectories, smoothing (1€ filter), velocity, acceleration, joint angles, head pitch/roll/yaw, PCA; avatars.

Input

BlazePose CSV, Qualisys TSV, video

Output

kinematics as CSV, avatar videos, glb files

In Healthy Strings

Checking tracking quality next to the video and computing joint angles.

Watch out

Note the smoothing settings in your methods – they change velocity and acceleration.

The 2D counterpart: visualises and edits MoveNet, BlazePose and other 2D tracking data.

Input

2D tracking CSV, video

Output

kinematics and cleaned data (CSV)

In Healthy Strings

Cleaning 2D data from the 30R video.

3D Pose Tracking Visualiser (Avatar) ↗

Shows pose-tracking results as an avatar next to the original video.

Input

BlazePose CSV, video

Output

avatar video

In Healthy Strings

Showing results to musicians and teachers.

Stick figure of the tracked pose, synchronised with the video.

Input

BlazePose CSV, video

Output

stick-figure video

In Healthy Strings

Spotting tracking errors quickly.

Excel Synchroniser ↗

Brings tables with different sampling rates onto one time axis.

Input

CSV/Excel tables with time columns

Output

one merged table

In Healthy Strings

Merging pose data (60 fps), audio features and ratings.

Watch out

Check the time offset between cameras first (clap).

Moving average with an adjustable window.

Input

CSV/Excel

Output

smoothed columns

In Healthy Strings

Smoothing noisy landmark traces.

Watch out

Smoothing hides fast movements such as vibrato.

z-standardises every column.

Input

CSV/Excel

Output

z-values

In Healthy Strings

Comparing players with different body size or recordings with different levels.

Converts between Excel and CSV formats (separators, decimal comma or point).

Input

CSV/Excel

Output

converted file

In Healthy Strings

Fixing the decimal comma of German-language exports before R or Python.

Finds correlations between columns.

Input

CSV/Excel

Output

correlation matrix

In Healthy Strings

First look at relations between movement and sound features.

Watch out

Correlation is not causation; time series are autocorrelated.

Tests differences between columns.

Input

CSV/Excel

Output

t-tests

In Healthy Strings

Simple comparisons, e.g. two conditions.

Watch out

Many tests → correct for multiple testing.

Fleiss' kappa for agreement between raters.

Input

CSV/Excel with ratings

Output

kappa values

In Healthy Strings

Agreement of physiotherapists or teachers on categories.

Watch out

For 0–100 scales use the ICC (R).

ANOVA and PCA to find patterns in multidimensional data.

Input

CSV/Excel

Output

ANOVA tables, PCA

In Healthy Strings

Exploring differences between outfits, tasks or players.

Analyses time series such as audio, physiological or movement data.

Input

CSV/Excel

Output

time-series statistics

In Healthy Strings

Looking at a whole performance curve.

Charts and plots from Excel or CSV tables.

Input

CSV/Excel

Output

charts

In Healthy Strings

Quick figures for meetings.

Box, violin or radar plots.

Input

CSV/Excel

Output

plots

In Healthy Strings

Distributions per outfit, task or group.

Heatmaps from Excel or CSV tables.

Input

CSV/Excel

Output

heatmap

In Healthy Strings

Overview of many points × tasks.