[{"id":4393,"date":"2025-07-25T14:53:36","date_gmt":"2025-07-25T14:53:36","guid":{"rendered":"https:\/\/edspace.american.edu\/openbehavior\/?p=4393"},"modified":"2025-07-25T14:53:38","modified_gmt":"2025-07-25T14:53:38","slug":"video-methods-update-summer-2025","status":"publish","type":"post","link":"https:\/\/edspace.american.edu\/openbehavior\/2025\/07\/25\/video-methods-update-summer-2025\/","title":{"rendered":"Video Methods Update: Summer 2025"},"content":{"rendered":"\n

The OpenBehavior team explored the use of three advanced video analysis tools in behavioral neuroscience during summer 2025. These tools are SLEAP<\/strong><\/a> for pose estimation, DeepEthogram<\/strong><\/a> for tracking and timing action sequences, and A-SOID<\/strong><\/a> for measuring movement dynamics from pose data. Our experience has been that these tools work very well together for carrying out comprehensive analyses of video data, in our case from experiments involving rodents performing operant tasks while freely moving.<\/p>\n\n\n\n

The team tested these tools using videos from the OpenBehavior Video Repository and ongoing research projects in the Laubach Lab at American University. The following instructions will help other researchers integrate these tools into their workflows.<\/p>\n\n\n\n

Note that these instructions are specific to PCs running the Linux Mint<\/strong><\/a> operating system. We highly recommend setting up PCs with Mint or other Linux distributions, as Linux provides a user-friendly platform for scientific data analysis, especially with GPU computing.<\/p>\n\n\n\n

These instructions were written by Jason Blackmer, an undergraduate neuroscience and data science major at American University, and Dr. Mark Laubach, a neuroscience professor at American University. Image credit: Samantha R. White<\/em><\/p>\n\n\n\n

\n\n\n\n

Python Environment Setup<\/h2>\n\n\n\n

We prefer Mamba<\/strong> over Conda as a package manager for Python due to its superior speed. Detailed documentation for Mamba can be found here: https:\/\/mamba.readthedocs.io\/en\/latest\/user_guide\/mamba.html<\/a><\/p>\n\n\n\n

To begin, install a minimal Python environment using the Miniforge distribution, following the instructions on this page: https:\/\/github.com\/conda-forge\/miniforge<\/a><\/p>\n\n\n\n

Once Miniforge is installed, open a terminal and create a base environment using Python 3.12 (as many scientific packages are not compatible with Python 3.13 or 3.14):<\/p>\n\n\n\n

mamba create -n my_python_env python=3.12\nmamba activate my_python_env\n<\/code><\/pre>\n\n\n\n
\n\n\n\n
SLEAP Setup<\/h2>\n\n\n\n
The following instructions have enabled reproducible installations of SLEAP on Linux PCs with GPUs within our research group.<\/p>\n\n\n\n
First, configure Mamba channels:<\/p>\n\n\n\n
mamba config --add channels nvidia\nmamba config --add channels conda-forge\n<\/code><\/pre>\n\n\n\n
Next, update Mamba:<\/p>\n\n\n\n
mamba update -n base -c conda-forge mamba\n<\/code><\/pre>\n\n\n\n
Create and activate a new Mamba environment for SLEAP with Python 3.7.12:<\/p>\n\n\n\n
mamba create -n \"sleap\" python=3.7.12\nmamba activate sleap\n<\/code><\/pre>\n\n\n\n
Install necessary dependencies:<\/p>\n\n\n\n
mamba install scipy=1.4.1\nmamba install ipython\nmamba install cudatoolkit=11.3.1\nmamba install cudnn=8.2.1\nmamba install -c nvidia cuda-nvcc=11.3\nmamba install tensorflow=2.7.0\n<\/code><\/pre>\n\n\n\n
You should test GPU utilization using the example provided here: https:\/\/www.tensorflow.org\/tutorials\/quickstart\/beginner<\/a><\/p>\n\n\n\n
You can monitor GPU usage with nvtop, as described here https:\/\/github.com\/Syllo\/nvtop<\/a>.<\/p>\n\n\n\n
Finally, install SLEAP via pip:<\/p>\n\n\n\n
pip install sleap[pypi]\n<\/code><\/pre>\n\n\n\n
\n\n\n\n
DeepEthogram Setup<\/h2>\n\n\n\n
The installation of DeepEthogram presented several challenges. We update the project’s installation instructions and note some required modifications to the source code to achieve a successful setup.<\/p>\n\n\n\n
To begin, create and activate the DeepEthogram environment:<\/p>\n\n\n\n
mamba create -n deg python=3.8\nmamba activate deg\n<\/code><\/pre>\n\n\n\n
Install core dependencies:<\/p>\n\n\n\n
mamba install pytorch==1.11.0 torchvision==0.12.0 torchaudio==0.11.0 cudatoolkit=11.3 -c pytorch\nmamba install ipython\nmamba install pytorch-lightning=1.6.5\nmamba install tensorflow=2.11.1\n<\/code><\/pre>\n\n\n\n
Then, install DeepEthogram:<\/p>\n\n\n\n
pip install deepethogram\n<\/code><\/pre>\n\n\n\n
The GUI should now launch successfully:<\/p>\n\n\n\n
python -m deepethogram\n<\/code><\/pre>\n\n\n\n
A crucial step for efficient training involves video conversion. MP4 files, while space-efficient, result in extremely slow training times. Therefore, it is highly recommended to convert videos for DeepEthogram to a directory of either JPGs or PNGs.<\/p>\n\n\n\n
To convert videos, open a terminal in the folder containing all intended videos and launch an IPython session by typing python<\/code>:<\/p>\n\n\n\n
from deepethogram.projects import convert_all_videos\nconvert_all_videos('PATH\/TO\/MY\/project_config.yaml', movie_format='directory', codec = '.jpg')\n<\/code><\/pre>\n\n\n\n
Additionally, specific changes were necessary within the DeepEthogram source code for proper functionality. The team set 'val' : False<\/code> in line 129 of base.py<\/code>.<\/p>\n\n\n\n
Furthermore, because JPGs were used for video input, an extra argument had to be added within the VideoReader<\/code> call of the __getitem__<\/code> function in deepethogram\/datasets.py<\/code>. The original line:<\/p>\n\n\n\n
with VideoReader(self.videofile, assume_writer_style=True) as reader:\n<\/code><\/pre>\n\n\n\n
was modified to:<\/p>\n\n\n\n
with VideoReader(self.videofile, assume_writer_style=True, filetype='.jpg') as reader:\n<\/code><\/pre>\n\n\n\n
With these adjustments, DeepEthogram now runs smoothly for the team.<\/p>\n\n\n\n
\n\n\n\n
A-SOID Setup<\/h2>\n\n\n\n
We followed the setup guide provided on the A-SOID project’s GitHub page. However, for images to display correctly during one of the steps, OpenCV needed to be uninstalled from pip and then rebuilt using Mamba.<\/p>\n\n\n\n
Therefore, after installing and setting up the A-SOID environment, the following commands were executed:<\/p>\n\n\n\n
pip uninstall opencv-python\nmamba install -c conda-forge opencv\n<\/code><\/pre>\n\n\n\n
Furthermore, a few crucial changes were required to ensure compatibility and smooth operation of DeepEthogram and SLEAP with A-SOID. A ‘time’ column was added to the DeepEthogram label data:<\/p>\n\n\n\n
import numpy as np\nimport pandas as pd\ndf = pd.read_csv('PATH\/TO\/labels.csv', index_col= 0)\ndf['time'] = df.index*0.04 # This is just 1\/framerate, our labels were on 25fps.\ndf.to_csv(path_or_buf='PATH\/TO\/labels.csv')\n```<\/code><\/pre>\n\n\n\n
Secondly, a significant issue arose in the A-SOID manual refinement phase. A function from the preprocessing script needed to be edited because the adp_filt<\/code> function was narrowing the pose data to the selected keypoints but not similarly processing the likelihood (llh) values. This led to mismatched DataFrame dimensions if not all keypoints were selected, causing the operation to fail. To rectify this, the following lines were added to the adp_filt<\/code> definition in A-SOID\/asoid\/utils\/preprocessing.py<\/code>:<\/p>\n\n\n\n
def adp_filt(pose, idx_selected, idx_llh, llh_value):\n    ### NEW LINE:\n    idx_llh_selected = [int(i)+2 for i in idx_selected[0::2]]\n\n    # Unaltered\n    datax = np.array(pose.iloc[:, idx_selected[::2]])\n    datay = np.array(pose.iloc[:, idx_selected[1::2]])\n\n    ### MODIFIED THE REFERENCE HERE:\n    data_lh = np.array(pose.iloc[:, idx_llh_selected])\n    # Formerly: data_lh = np.array(pose.iloc[:, idx_llh])\n<\/code><\/pre>\n\n\n\n
This modification completely resolves the issue, allowing both DeepEthogram and SLEAP to be used effectively with A-SOID.<\/p>\n","protected":false},"excerpt":{"rendered":"
The OpenBehavior team explored the use of three advanced video analysis tools in behavioral neuroscience during summer 2025. These tools are SLEAP for pose estimation, DeepEthogram for tracking and timing action sequences, and A-SOID for measuring movement dynamics from pose data. Our experience has been that these tools work very well together for carrying out […]<\/p>\n","protected":false},"author":2065,"featured_media":4397,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":true,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[1805,1804,1,2281],"tags":[67],"class_list":["post-4393","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-in-the-community","category-openbehavior-team-update","category-uncategorized","category-video-analysis","tag-open-source"],"jetpack_featured_media_url":"https:\/\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/07\/VideoRecordingAndAnalysis_Crop.jpg","jetpack-related-posts":[{"id":4279,"url":"https:\/\/edspace.american.edu\/openbehavior\/2025\/02\/13\/open-source-tools-for-data-analysis\/","url_meta":{"origin":4393,"position":0},"title":"Open-Source Tools for Data Analysis","author":"marklaubach","date":"February 13, 2025","format":false,"excerpt":"This year, we're focusing on the impact of open-source tools on research productivity, which seems especially relevant given current funding challenges for science. Our findings aim to demonstrate how collaborative, open efforts are increasing research capacity and efficiency within neuroscience. Low-cost solutions are crucial in this era of increasing financial\u2026","rel":"","context":"In "Community Call to Action"","block_context":{"text":"Community Call to Action","link":"https:\/\/edspace.american.edu\/openbehavior\/category\/community-call-to-action\/"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/02\/Screenshot-from-2025-02-12-15-52-17-1.png?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/02\/Screenshot-from-2025-02-12-15-52-17-1.png?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/02\/Screenshot-from-2025-02-12-15-52-17-1.png?resize=525%2C300&ssl=1 1.5x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/02\/Screenshot-from-2025-02-12-15-52-17-1.png?resize=700%2C400&ssl=1 2x"},"classes":[]},{"id":4346,"url":"https:\/\/edspace.american.edu\/openbehavior\/2025\/05\/16\/setups-and-protocols-video-analysis\/","url_meta":{"origin":4393,"position":1},"title":"Setups and Protocols: Video Analysis","author":"marklaubach","date":"May 16, 2025","format":false,"excerpt":"This summer, the OpenBehavior project team is focusing on creating documentation for setting up and using three valuable video analysis tools in neuroscience research. These tools often lack straightforward, in-lab setup guides, so we want to make them more accessible. First, we will provide a step-by-step guide on how to\u2026","rel":"","context":"In "In the Community"","block_context":{"text":"In the Community","link":"https:\/\/edspace.american.edu\/openbehavior\/category\/in-the-community\/"},"img":{"alt_text":"Rat performing a decision making task","src":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/05\/Screenshot-from-2025-05-16-11-52-50.png?resize=350%2C200&ssl=1","width":350,"height":200},"classes":[]},{"id":3058,"url":"https:\/\/edspace.american.edu\/openbehavior\/2022\/04\/07\/video-methods-setup-standards\/","url_meta":{"origin":4393,"position":2},"title":"Video methods: Setup and standards","author":"marklaubach","date":"April 7, 2022","format":false,"excerpt":"In the summer of 2021, OpenBehavior organized a working group on methods for video analysis. A Slack channel was started to discuss the methods and a series of virtual meetings were held that involved conversations among users and developers. Some in the group decided to write a commentary on the\u2026","rel":"","context":"In "Community Call to Action"","block_context":{"text":"Community Call to Action","link":"https:\/\/edspace.american.edu\/openbehavior\/category\/community-call-to-action\/"},"img":{"alt_text":"Video Setups","src":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2022\/04\/Figure-1.png?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2022\/04\/Figure-1.png?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2022\/04\/Figure-1.png?resize=525%2C300&ssl=1 1.5x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2022\/04\/Figure-1.png?resize=700%2C400&ssl=1 2x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2022\/04\/Figure-1.png?resize=1050%2C600&ssl=1 3x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2022\/04\/Figure-1.png?resize=1400%2C800&ssl=1 4x"},"classes":[]},{"id":2790,"url":"https:\/\/edspace.american.edu\/openbehavior\/2021\/06\/17\/community-conversations-video-analysis\/","url_meta":{"origin":4393,"position":3},"title":"Community Conversations on Video Analysis","author":"marklaubach","date":"June 17, 2021","format":false,"excerpt":"Following several conversations on Twitter about methods for video analysis, in May 2021 the OpenBehavior team reached out to prominent research groups working on methods for video analysis and groups who contributed videos to our behavioral neuroscience video repository to organize what we called Community Conversations on Video Analysis. Two\u2026","rel":"","context":"In "Community Call to Action"","block_context":{"text":"Community Call to Action","link":"https:\/\/edspace.american.edu\/openbehavior\/category\/community-call-to-action\/"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior24\/wp-content\/uploads\/sites\/2431\/2021\/06\/filename00075-1024x576.png?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior24\/wp-content\/uploads\/sites\/2431\/2021\/06\/filename00075-1024x576.png?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior24\/wp-content\/uploads\/sites\/2431\/2021\/06\/filename00075-1024x576.png?resize=525%2C300&ssl=1 1.5x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior24\/wp-content\/uploads\/sites\/2431\/2021\/06\/filename00075-1024x576.png?resize=700%2C400&ssl=1 2x"},"classes":[]},{"id":4391,"url":"https:\/\/edspace.american.edu\/openbehavior\/2025\/07\/18\/impactful-open-source-projects-in-2025\/","url_meta":{"origin":4393,"position":4},"title":"Impactful Open-Source Projects in 2025","author":"marklaubach","date":"July 18, 2025","format":false,"excerpt":"Over the past several months, we've highlighted open-source projects that have captured the interest of our followers, focusing on specific categories such as behavioral devices or video analysis programs. In this final post of the series, we review projects that have generated significant interest during the first half of 2025.\u2026","rel":"","context":"In "In the Community"","block_context":{"text":"In the Community","link":"https:\/\/edspace.american.edu\/openbehavior\/category\/in-the-community\/"},"img":{"alt_text":"word cloud of page hits on the OB blog in the first half of 2025","src":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/07\/OB-Hits-2025.png?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/07\/OB-Hits-2025.png?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/07\/OB-Hits-2025.png?resize=525%2C300&ssl=1 1.5x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/07\/OB-Hits-2025.png?resize=700%2C400&ssl=1 2x"},"classes":[]},{"id":4357,"url":"https:\/\/edspace.american.edu\/openbehavior\/2025\/06\/06\/devices-and-programs-for-electrophysiology\/","url_meta":{"origin":4393,"position":5},"title":"Devices and Programs for Electrophysiology","author":"marklaubach","date":"June 6, 2025","format":false,"excerpt":"The OpenBehavior Project primarily focuses on tools for behavioral testing and data analysis. However, our database also features a growing number of valuable open-source tools for electrophysiology experiments that have been reported on our website over the past five years. This week's post summarizes these key projects. The word cloud\u2026","rel":"","context":"In "In the Community"","block_context":{"text":"In the Community","link":"https:\/\/edspace.american.edu\/openbehavior\/category\/in-the-community\/"},"img":{"alt_text":"Word Cloud Based on Page Hits for Electrophysiology Projects","src":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/06\/OB-Ephys-2024.png?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/06\/OB-Ephys-2024.png?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/06\/OB-Ephys-2024.png?resize=525%2C300&ssl=1 1.5x, https:\/\/i0.wp.com\/edspace.american.edu\/openbehavior\/wp-content\/uploads\/sites\/2431\/2025\/06\/OB-Ephys-2024.png?resize=700%2C400&ssl=1 2x"},"classes":[]}],"jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/edspace.american.edu\/openbehavior\/wp-json\/wp\/v2\/posts\/4393","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/edspace.american.edu\/openbehavior\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/edspace.american.edu\/openbehavior\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/edspace.american.edu\/openbehavior\/wp-json\/wp\/v2\/users\/2065"}],"replies":[{"embeddable":true,"href":"https:\/\/edspace.american.edu\/openbehavior\/wp-json\/wp\/v2\/comments?post=4393"}],"version-history":[{"count":0,"href":"https:\/\/edspace.american.edu\/openbehavior\/wp-json\/wp\/v2\/posts\/4393\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/edspace.american.edu\/openbehavior\/wp-json\/wp\/v2\/media\/4397"}],"wp:attachment":[{"href":"https:\/\/edspace.american.edu\/openbehavior\/wp-json\/wp\/v2\/media?parent=4393"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/edspace.american.edu\/openbehavior\/wp-json\/wp\/v2\/categories?post=4393"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/edspace.american.edu\/openbehavior\/wp-json\/wp\/v2\/tags?post=4393"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}},{"id":4391,"date":"2025-07-18T14:30:22","date_gmt":"2025-07-18T14:30:22","guid":{"rendered":"https:\/\/edspace.american.edu\/openbehavior\/?p=4391"},"modified":"2025-07-18T14:30:24","modified_gmt":"2025-07-18T14:30:24","slug":"impactful-open-source-projects-in-2025","status":"publish","type":"post","link":"https:\/\/edspace.american.edu\/openbehavior\/2025\/07\/18\/impactful-open-source-projects-in-2025\/","title":{"rendered":"Impactful Open-Source Projects in 2025"},"content":{"rendered":"\n
Over the past several months, we’ve highlighted open-source projects that have captured the interest of our followers, focusing on specific categories such as behavioral devices or video analysis programs. In this final post of the series, we review projects that have generated significant interest during the first half of 2025.<\/p>\n\n\n\n
By analyzing page hits on our WordPress blog, we’ve observed a general decrease in overall page views this year. One potential reason for this trend is that our primary referral source, Twitter\/X, has seen a decline in usage among many neuroscientists over the past year. While we made efforts to generate interest for the OpenBehavior project on an alternative platform, Bluesky, traffic directed from that platform has been minimal. We believe this may be partly due to the limited presence or activity of many trainees and early-stage investigators on Bluesky, who, based on user feedback and attendance at our workshops and social events (like the open-source tools workshop at the SfN meeting), have historically been our most active followers.<\/p>\n\n\n\n
Nevertheless, the word cloud above represents the top thirty tools based on website traffic in the first half of 2025. These projects offer a diverse range of useful tools, including those for electrophysiology data analysis (ghostipy), histological quantification (Bell Jar), video analysis (BORIS), neural implants (DREAM), and behavioral apparatuses (NC4gate). Detailed descriptions of the top ten projects are provided below. We hope these projects prove valuable to the community.<\/p>\n\n\n\n
<\/p>\n\n\n\n
\n\n\n\n
    \n
  1. ghostipy<\/strong>\n