Related Experiment Videos
The Cascade Neural Network Model and a Speed-Accuracy Trade-Off of Arm Movement
M. Hirayama1, M. Kawato, M. I. Jordan
1ATR Human Information Processing Research Laboratories, 2ndas;2, Hikaridai Seika-cho, Soraku-gun, Kyoto 619ndash02, Japan.
Journal of Motor Behavior
|September 1, 1993
Summary
A new hybrid neural network model simulates aimed arm movements. This model explains time-accuracy trade-offs and movement variability in motor control, offering insights into neural mechanisms.
Area of Science:
- Neuroscience
- Robotics
- Computational Motor Control
Background:
- Aimed arm movements involve complex motor control.
- Existing models may not fully capture the nuances of motor command generation.
- Understanding the neural basis of movement variability is crucial.
Purpose of the Study:
- To propose a hybrid neural network model for aimed arm movements.
- To investigate the computational mechanisms underlying motor control and movement variability.
- To account for invariant features of multijoint arm trajectories.
Main Methods:
- A hybrid neural network model combining feedforward and postural controllers.
- Utilizing the cascade neural network (Kawato et al., 1990) for feedforward control.
- Simulating target-directed arm movements with specific parameter constraints.
Main Results:
- The model generated a planning time-accuracy trade-off.
- A quasi-power-law speed-accuracy trade-off was observed.
- The model accounts for stochastic variability in motor commands and invariant trajectory features.
Conclusions:
- The proposed hybrid model offers a candidate neural mechanism for aimed arm movements.
- The model explains observed trade-offs and variability in motor control.
- It successfully replicates key characteristics of human arm trajectories.