Positions leave out shape
A 2D position can show where an animal is, while leaving out how its body is posed and moving through space.
A clearer view of animal behavior
See movement
in a new dimension.
Changes in the way an animal moves can reveal changes in health and well-being. Behavini is developing AI-assisted 3D body modelling to capture a richer, more useful picture of that movement.
Get in touchWHY IT MATTERS
Behavior is more than a point on a screen. Shape, appearance, and subtle shifts can tell a deeper story.
The problem
Small changes are
easy to miss.
One mouse pose, shown as a schematic position, keypoint map, and body model. These illustrations are not model output.
Keypoints tell part of the story
Sparse landmarks are useful, but may miss subtle changes in body shape, posture, and motion between tracked points.
Richer measurements can help
Explicit 3D body models could make those changes easier to quantify, compare, and interpret over time.
Our approach
From pixels
to understanding.
We combine explicit 3D body modelling with AI to work toward movement measurements that researchers can inspect and use.
Where it could help
New ways to ask
better questions.
Encouraged by evidence in the literature, we anticipate substantial advances in these areas through our approach.
Earlier disease signals
Study prodromal and early disease-related changes in animal models through more detailed movement measures.
Drug effect timelines
Measure behavioral effects and explore when meaningful changes begin to emerge.
Synthetic data
Generate data to support model training and evaluation across movement scenarios and species.
What we are building
Built for a
richer picture.
Behavini is in development. These areas describe our design goals across species, based on our findings in monkeys and humans.
Development goalsBody shape & movement
More realistic representations of how bodies are shaped, posed, and moving in 3D.
Interpretable analysis
Explainable AI and movement measurements researchers can examine and relate back to behavior.
Across species, over time
Adaptation to different species in research and industry through developing species-specific models and synthetic pretraining.
Research
Evidence belongs
in the open.
BigMaQ: A Big Macaque Motion and Animation Dataset Bridging Image and 3D Pose Representations
Realistic monkey body animation reveals an uncanny valley in macaque body perception
Digital Gait Measures Capture 1-Year Progression in Early-Stage Spinocerebellar Ataxia Type 2
Real-life gait assessment in degenerative cerebellar ataxia: Toward ecologically valid biomarkers
These publications do not constitute validation of Behavini’s technology.
The people
Curious about
how bodies move.
Research experience spanning 3D animal motion and real-world gait measurement.
3D motion & computer vision
Lucas M. Martini1,2
Researches computational neuroscience and computer vision, with a focus on 3D pose reconstruction, motion capture, and animal body modelling.
Institutional profile
Digital gait measurement
Jens Seemann1,2
Studies real-world gait in neurological disease, using movement measurements to examine subtle changes and progression in ataxia.
Institutional profileResearch affiliations
- University of Tübingen
- International Max Planck Research School for Intelligent Systems (IMPRS-IS)
Get in touch
Let's make sense
of movement.
We welcome conversations with research collaborators, pilot partners, and funding partners interested in better tools for animal behavioral analysis.
EMAIL UScontact@behavini.ai