Tutorials

• Interacting with NeuroMechFly
  • Markov Decision Process (MDP) and the Gym API
  • Example: Kinematic replay of experimentally recorded behavior
• Controlling locomotion with CPGs
  • The CPG network
  • Controlling leg stepping with CPGs
  • Plugging the controller into the simulation
  • Leg adhesion
• Rule-based controller
  • Preprogrammed steps, refactored
  • Implementing the rules
  • Plugging the controller into the simulation
• Overcome complex terrain with a hybrid controller
  • Defining the arena
  • Basic CPG- and rule-based controllers
  • Building a hybrid controller
• Encapsulating custom code into Gym environments: A turning controller
  • Gym environments and MDP
  • Approach for turning
  • Implementing the HybridTurningController class
• Vision basics
  • Retina simulation
  • A dynamic arena with a moving sphere
  • Visual feature preprocessing
  • Implementing a object tracking controller
• Olfaction basics
  • Odor arena
  • Controller for odor taxis
• Path integration
  • Introduction
  • The algorithm
  • Collecting walking data
  • Extracting input and target variables from simulation data
  • Training models to predict changes in locomotor state
  • Integrating changes in locomotor state to estimate position
• Head stabilization
  • Introduction
  • Collecting training data
  • Training an internal model to control neck actuation
  • Deploying the model
• Connectome-constrained visual system model
  • Simulation with multiple flies
  • Interfacing NeuroMechFly with a connectome-constrained vision model
• Advanced olfaction: Navigating a complex odor plume
  • Simulating a complex odor plume
  • Creating a plume within an arena
  • Plume tracking task
  • Implementing a plume tracking controller