Discover Hidden Behavior Patterns
Harness AI to Identify Patterns Without Labels.
Compatible with any recording system—capture every motion from any angle.
Translate motion into detailed datasets.
Behavioral Decomposition
Our software uses an autoregressive hidden Markov model (AR-HMM) to identify recurring motifs of behavior, called ‘syllables,’ and their transitions.
Raw Movement Data
Syllables
Transitions
Dimensionality Reduction
Raw motion data is simplified using PCA (Principal Component Analysis) to keep key features while reducing noise and complexity.
Behavioral Segmentation
The AR-HMM breaks behavior into sub-second “syllables,” representing repeated patterns like “dart,” “pause,” or “groom.”
Behavioral Grammar
The system maps how these syllables transition over time—creating a unique grammar of behavior for each subject.
This process reveals hidden structure in behavior and provides rich, reproducible insights for neuroscience research.
Applications
Drug Screening
Behavioral Science
Neurological Research
Ecological Studies
Why it Matters
Versatile
Camera cycling through different setups.
Accurate Models
Markov chain diagram highlighting transitions.
Scalable Insights
Cluster of bar graphs growing to represent large datasets.
