A new algorithm and system for the extraction of delta-lognormal parameters

In this report, we present a new analytical method to estimate the parameters of Delta-lognormal functions. According to the Kinematic Theory of rapid human movements, these parameters contain information both on the motor commands and on the timing properties of a neuromuscular system. The new algorithm, called XZERO, exploits relationships between the zero crossings of the first and the second time-derivatives of a lognormal function and its four basic parameters. The methodology is described and evaluated in various testing conditions. Furthermore, for the first time, the extraction accuracy is quantified empirically, taking advantage of the exponential relationships that link the disperssion of the extraction errors with its signal to noise ratio. A new extraction system, which uses a benchmark of three estimation methods, is also proposed and evaluated in the mid-term perspective of developing machine intelligence applications that rely on lognormal functions.

Mots clés

• Pattern recognition
• Kinematic Theory
• rapid movement
• Delta-Lognormal model
• lognormal function
• parameter extraction
• motor control
• nonlinear regression
• optimization
• curve fitting.