2D Data Analyzer

1. Pose Data (BP/MN):
1b. optional: Hands (MediaPipe):
1c. optional: FaceTracking:
1d. optional: PointTracking:
2. optional: Load Video:
optional: save edited motion data
Wait for Data ...
Frame: 0 / 0
2D Data Analyzer © 2026 by Christoph Reuter and SInES (last modified: 6.10.2026) is released under MIT License. This tool includes movement features according to the RITMO VideoAnalysis (Jensenius 2005), Matlab MoCap Toolbox (Burger, Toiviainen 2013), PyEyesWeb (Sabharwal, Corbellini, Ghisio, Coletta, Romano, Al Foysal, Volpe & Camurri 2026) and to the PCA Movement Synergy Analysis (Bigand, Prigent, Berret, Braffort 2021)

How to cite: Reuter, C. (2026). SInES Visualizers: 2D Data Analyzer. Web Application https://sinestools.univie.ac.at/movenet_visualizer.htm



AI & Rigging Tools

Exports perfect skeleton maps for Stable Diffusion & MagicAnimate.
0 Frames selected

Exported Movement Features (2D Dictionary)

Feature Name Range / Unit Description
Dist_kumuliert[0, ∞) pxTotal distance traveled (cumulative) by the center of mass or the respective marker.
Speed / Acc / Jerk[0, ∞)Velocity (px/s), acceleration (px/s²), and jerk (px/s³) represent the first, second, and third mathematical derivatives of position over time.
FluidityRatio [0, ∞)The ratio of speed to acceleration. Higher values indicate smoother motion (fewer abrupt changes in speed).
Flux[0, ∞)Kinematic energy flux (Speed * Acceleration). A combined measure of movement intensity.
Spread[0, ∞) pxSpatial distribution. The average distance of all body markers from the current common center of mass (centroid).
BoundingBox_Area[0, ∞) px²Area of the smallest 2D rectangle (width * height) that encloses the entire body posture on screen in this frame.
KinEnergy_Win[0, ∞) proxyProxy for kinetic energy (velocity²) averaged over the defined sliding window (e.g., 0.5 s).
Smoothness_Win[0, 1]A measure of the “smoothness” of motion, based on jerk. 1.0 = perfectly smooth, ~0.0 = extremely jerky/abrupt.
Contraction_Win[0, ∞) px²Average bounding box area over the time window. Shows the extent to which the body contracts or expands on screen over time.
DirChange_Win[0, ∞) DegAccumulated changes in direction (angles in degrees) of the motion vectors over the time window. Shows a zigzag pattern.
EllipseSphericity[0, 1]Describes the spatial distribution of the markers (2D covariance matrix). 1.0 = perfectly circular posture; close to 0.0 = extremely flat or elongated posture.
PointsDensity[0, ∞) 1/px²Number of markers divided by the bounding box area. Indicates the visual compactness of the pose.
Equilibrium_Offset[0, ∞) pxHorizontal distance (X-axis) between the center of gravity and the center of the ankles (base of support). An indicator of physical balance and forward/backward lean.
Suddenness_Win[0, ∞)The maximum jerk within the time window. Highlights sudden, unexpected surges.
Impulsivity_Win[0, ∞)Maximum acceleration within the window, scaled by the difference between the maximum and minimum energy (amplitude).
Lightness_Win[-1, 1]Ratio of vertical velocity to total velocity. Positive values = upward motion (light), negative values = downward motion (heavy).
InterPerson_Distance[0, ∞) px(Multi-person only) The absolute distance between the 2D centers of mass of two people on screen.
InterPerson_SpreadSymmetry[0, 1](Multi-person only) Compares the “spread” (space occupied) of two people. 1.0 = exactly the same, 0.0 = extremely asymmetrical.
InterPerson_MovementSynchrony[-1, 1](Multi-person only) Cosine similarity of the motion vectors (centers). 1.0 = moving in exactly the same direction; -1.0 = moving in opposite directions.
Symmetry_Distance[0, 1]Measures whether the left and right joints (e.g., wrists) are the same distance from the center. 1.0 = perfect symmetry.
Symmetry_Direction[-1, 1]Cosine similarity of the velocity vectors of the left and right joint pairs. 1.0 = moving synchronously in the same direction.