Research‎ > ‎Seminars‎ > ‎Seminars 2010‎ > ‎

2010.04.21 - NCM: Day 1

  • Finger tip moisture is optimally modulated during object manipulation
    • moisture always seemed to approach a constant value
    • Grip force decreased over trials
  • Tricking CNS into believing in movement
    • Feeling holes and bumps
      • just knowing lateral force is ambiguous need know relative motion
      • lateral force wins in sensory conflict
    • Illusions
      • out-of-phase array of pinching-stretching feels like movement
      • moving one hand and pressure cues to the other produces a combined illusion
      • transference of motion between vision and touch by using moving visual and sensory fields
  • Alan Smith - Slip signals in touch
    • Touch consists of the skin "sticking-slipping"
    • Experiments show normal force decreases with increasing roughness (what about surface roughness?)
    • Is there a slip signal? Maybe this is actually just a desire of the nervous system to equalize tangential stress across surface of skin
    • Preferred level of friction about 0.6-0.7
  • Jerry Loeb - Biomimetic robot finger
    • measuring impedance of a deformable balloon
    • requires database of known objects
  • Atsushi Nambu - Cortico-basal ganglia loops and disorders
    • Hyperdirect - excitation
      • Turn brake on
    • Direct - inhibition
      • Release only desired motor program
    • Indirect - excitation
      • Turn brake on
  • Watanabe - Conflict resolution in basal ganglia
    • Automatic and volitional commands compete (stretch reflex vs APR)
    • Without the basal ganglia you need significantly more activity to accurately resolve conflict but with the basal ganglia you only need one-to-one power of automatic and volitional commands
  • Peter Redgrave - Does phasic dopamine signaling reinforce learning?
    • Dopamine as a reward was found to predict reward, but looking at the timing they must be able to do this before motor output
    • So where do the dopamine inputs come from?
      • Eyes->superior colliculus->substantia nigra->Dopamine
      • Superior colliculus is likely the input for visually evoked dopamine response
    • Not used to generate error signal on reward prediction
    • Likely used to test if we caused the event or if the event just occured
    • Fighting against Wolfram
  • Ann Graybiel - Habit formation
    • Once a habit is formed it sticks
    • Striatum is a learning machine
  • Stefan Schaal - Modularity adaptation and learning with dynamic movement primitives
    • Path integral reinforcement learning
  • Jochen Steil - Toward intelligent robots
    • Compliant robots
    • Modular architecture
  • Terrence Sanger - Stochastic control using linear operators on probability space
    • Fancy method to come up with stuff that looks like synergies
  • Jeroen Smeets - Can synergies during movements be determined from the structure of variability?
    • No they cannot.
  • Paul Chisek - The mechanism of compensation of intersegmental dynamics includes cortical control of bi-articular muscles
    • TMS perturbation to screw up cortical stuff
    • Shocking the motor cortex appears that it preferentially selects bi-articular muscles
Comments