Discrimination of Motion Direction in a Robot Using a Phenomenological Model of Synaptic Plasticity

Area of Science:

• Computational Neuroscience
• Robotics
• Neuroethology

Background:

• Directional motion detection is vital for animal behavior.
• Primary visual cortex (V1) exhibits direction selectivity in response to visual stimuli.
• Short-term synaptic plasticity (STP) is a known neural mechanism.

Purpose of the Study:

• To investigate if a bio-inspired model of STP in a robotic agent can discriminate real-world motion direction.
• To develop a spiking neural microcircuit that mimics V1's direction-selective properties.
• To validate computational models using physical robotic implementations.

Main Methods:

• Implemented a well-established short-term synaptic plasticity (STP) model in a spiking neural microcircuit.
• Designed a robotic agent to process sensory inputs and interact with real-world stimuli.
• Compared model responses to multielectrode recordings from macaque monkey V1 using sinusoidal gratings.

Main Results:

• The STP mechanism enabled direction-dependent synaptic changes in the microcircuit.
• The model successfully generated direction selectivity, a key V1 response property.
• Robotic agent responses correlated with experimental data from primate V1.

Conclusions:

• Short-term synaptic plasticity (STP) plays a significant role in explaining neural direction selectivity.
• A bio-inspired robotic model can effectively validate neurophysiological findings.
• This approach bridges computational modeling and experimental neuroscience for understanding motion perception.