Behavioral analysis is performed on animals such as Octopus, Monkey and Mice using transfer learning tools like DeepLabCut and steaming dimension reductions tools like proSVD.
# download the repo
git clone https://github.com/draeloslab/stim_behavior
cd stim_behavior
# install dependencies
conda env create --file=environment.yml
# install the repo locally
conda activate stim_behavior
pip install -e .
- Link to the Paper -> https://pubmed.ncbi.nlm.nih.gov/31551604/
- Download (raw) data from -> https://labshare.cshl.edu/shares/library/repository/38599/2pData/Animals/
- The scripts required to perform the following are inside
scripts/musall-mice/
. - The raw data contains the SVD components of the video data. Use
download_data.py
to download files more easily. - Once the files are downloaded to the local system, the video svd files needs to be combined to restore the original video. Run
construct_video.py
Applying DeepLabCut for real time hand tracking in collaboration with the Chestek Lab.
- Scripts
-
- cameraCtrl.py and camera.py: current camera interfaces for recording and accessed Imaging Source camera frames during NHP experiments. Use python cameraCtrl.py -h for help
-
- dataLoader.py: Class used to access DLC models for training and refining. dlcLiveTesting is an example notebook that walks through how to use DLC with exisiting NHP data. DataLoader class is used to preprocess and analyze DLC and SLEAP csv files. Use dmpFatigueFigure for an example notebook walking through the class.
Tree shrew with cranial window over V1, with Gcamp8s injected. Eye tracking camera records face and eye. 2P recording of neurons in V1. Original experiment and data collection by: Madineh Sarvestani @ MPFI/Cornell.
- Download the data from this dropbox link here: Link.
- Unzip and move it to a folder of your choice
- Run
python scripts/sarvestani_treeshrew/main.py
to extract the flourescence signals from the raw data using the suite2p package. - Suffix
-h
to understand command-line args.
- Download sample videos of the octopus experiement conducted in Dr.Pelled lab (MSU) from this dropbox link here: Link.
- Execute
scripts/pelled_octopus/dlc_compute_octopus.py
to perform deeplabcut analysis on the octopus videos. - Execute
scripts/pelled_octopus/prosvd_compute_octopus.py
to perform prosvd analysis on the octopus videos.