Flynet neural network and GUI application for automatically extracting wing kinematics from movie sequences used in Melis, Siwanowicz, and Dickinson 2024. Contains code for generating extended data figure 2.

System Requirements

Tested on ubuntu 22.04 with Python 3.11.4

Build Dependencies

The following libraries are required in order to build the software. They should be installed using apt prior to installing the software itself.

sudo apt install qtbase5-dev libvtk9-dev libarmadillo-dev libboost-python-dev libboost-numpy-dev

Dependencies

The complete list of software dependencies can be found in the "dependencies" section of the pyproject.toml file. All of the dependencies, aside from the build dependencies listed above, will be automatically installed by the Package Installer for Python (pip) when the software is installed.

Installation

Requires a working installation of Python 3.11.4. To install cd into the source directory and run

pip install .

The installation typically takes around 15-20min.

Running the flynet-app application

A set of example data for running the flynet GUI application are provided on CaltechData (TODO add location).

• Download the flynet.zip file from CaltechDATA here, unzip the archive and place the resulting flynet_data directory in the user's home directory.

• To start the flynet application run 'flynet-app' on the command line.

Training the flynet neural network

An example script, train_flynet.py, demonstrating how the flynet neural network is trained is provided in the 'examples' subdirectory. A set of example data for running the flynet GUI application are provided on CaltechData (TODO add location).

• Download the flynet.zip file from CaltechDATA here , unzip the archive and place the resulting flynet_data directory in the user's home directory.

• Run the train_fly.py script from within the 'examples' subdirectory.