Vision
Video analysis for behavioural research

Stoodia Vision

Turn experiment recordings and live camera streams into trajectories you can check, metrics you can trace and group comparisons ready for a report. The research task is defined by a plug-in scenario.

  • Videos stay on your computer
  • Scenario version is recorded
  • Every trajectory can be reviewed
upper zonelower zone
Schematic: a fish trajectory in an arena with analysis zones
Scenarios as plug-ins

One application, a scenario for each experiment

Stoodia Vision provides the project, video, review and statistics. A scenario adds the science for a specific type of recording.

01

Recording setup

Camera view, subject and arena layout: a tank seen from the side, from above or a multiwell plate.

02

Tracker and detector

Each scenario brings a tracking engine and a model trained for its kind of footage.

03

Metrics

The measurements that make sense for this setup, with quality-control indicators.

04

Reports and export

Visualizations, report layouts and export formats with the method and data provenance.

Scenarios are installed from the catalogue as signed packages. Each one has its own version and is updated without reinstalling the application.

How it works

The whole experiment in one project

  1. 1

    Set up the experiment

    Groups, conditions, arenas, analysis zones and physical scale.

  2. 2

    Add video

    Import recordings or record from one or several cameras.

  3. 3

    Track

    Works on CPU. An NVIDIA GPU with CUDA is recommended for fast tracking.

  4. 4

    Review trajectories

    Check gaps and false detections against the video, fix intervals, record the quality decision.

  5. 5

    Analyse behaviour

    Movement and spatial metrics, events, states and transitions over chosen time intervals.

  6. 6

    Compare and export

    Distributions, confidence intervals, effect sizes and statistical tests, exported with the calculation parameters.

Available scenarios

Choose the scenario that matches your recording

Install only what your experiment needs. Recordings are compared only within the same scenario.

Fish

ReadyVersion 1.4.0

Adult fish, side view

One adult fish per arena, filmed through the side wall of a tank.

Measures
  • distance travelled
  • mean speed
  • time in the top and bottom zones
PreviewVersion 1.4.0

Adult fish, top view

One adult fish per arena, filmed from above.

Measures
  • distance travelled
  • mean speed
  • time in the centre zone
PreviewVersion 1.7.0

Fish larvae in a multiwell plate

Head, body contour, centre and tail of one larva in every well. 24- and 96-well templates and custom grids.

Measures
  • distance per well
  • mean speed
  • path straightness
  • plate summary
Capabilities

Built for the laboratory workflow

  • Projects with several experimental groups
  • Video import and multi-camera recording
  • Local tracking on CPU or NVIDIA GPU
  • Batch processing of a whole project
  • Quality control and manual review
  • Spatial scale and unit calibration
  • Time-interval and event analysis
  • Group comparison and statistics
  • Export with parameters and provenance
  • Interface in 7 languages
Requirements

What you need to start

Operating system
Windows 10 or 11, 64-bit.
Compute
Works on CPU. An NVIDIA GPU with CUDA is recommended for fast tracking.
Data
Projects and source videos are stored locally. Cloud features are optional.
Download for WindowsStoodia Vision v.0.1.10 · Windows x64

The Windows installer is not yet signed with a publisher certificate. The application verifies the authenticity of updates and scenario packages.

Limits of use

Each scenario is validated only for its own recording conditions, and the application shows its status when you choose it. “Adult fish, side view” is ready for work; the other Fish scenarios are previews.

New recording conditions, other species and experiments in which animals occlude one another need visual review of trajectories and a hand-annotated validation sample of your own. Stoodia Vision provides the tools for that review but does not replace the researcher’s scientific judgement.