Home
Research
People
Publications
Tools
|
Motor control
of direction: a biologically inspired neural network model
MSc Thesis (pdf,
2.8 Mb)
At a general level, the research presented
in this thesis investigates how cognitive functions in the brain emerge
from the properties of basic components when these interact and function
cooperatively. The specific focus is on two topics: cortical control of
movement direction and visuomotor mapping of directional information.
For the visual guidance of movement to
be efficiently implemented in artificial systems, it is essential to understand
the computational mechanisms underpinning biological motor control. This
thesis explores the development of cortical control of movement direction
and visuomotor mapping within a biologically inspired computational framework
based on spike-processing neural networks. First, the implementation of
SpikeNNS simulator is described, followed by the detailed presentation
of two neural network models.
Chapter 1 represents an introduction
to the field of computational neuroscience and outlines the objectives
and the structure of the thesis. Chapter 2 provides a biological and modeling
framework of cortical control of motion direction. Chapter 3 introduces
a number of new theories on the biological and computational mechanisms
of perception-action coupling. Chapter 4 focuses upon the description of
the spiking neural model implemented. Chapter 5 presents the implementation
of the SpikeNNS simulator. Chapter 6 describes in turn, the model of motor
cortex self-organization and that of visuomotor mapping learning. Chapter
7 presents a discussion of the neurophysiological and theoretical implications
of the models.
Frontmatter (pdf)
1. Introduction
(pdf)
Part I. Neurobiology
and models of motor control of direction
2. Control
of movement direction (pdf)
3. Visuomotor
development (pdf)
Part II. SpikeNNS
- a simulator for spiking processing networks
4. Spiking
neural model (pdf)
5. Simulator
design and implementation (pdf)
Part III. Applications
6. Neural
networks architectures and results (pdf)
A. Self-organization
of neurons in motor cortex for coding the direction of movement (results)
B. Development
of visuomotor alignment of directional neural codes (results)
7. Discussion
and future directions (pdf)
References (pdf)
Funded by:
Basic Research programme
of Enterprise Ireland - 1999-2001.
Selected
references
|
