Resources
Overview
This page provides a curated collection of resources relevant to BBQS projects and working groups. You’ll find software tools and libraries for analysis, pre-trained ML models, multimodal datasets, evaluation benchmarks, academic publications, data platforms, and standards for consistent data sharing. Use the search below to filter resources by keywords, categories, target species, or working groups.
No matching resources found. Try a different search term.
General Tools
Software tools, libraries, or scripts used for analysis, visualization, tracking, etc.
idtracker.ai
Open-source multi-animal tracking software that can identify and track up to 100 unmarked animals in video, maintaining individual identities throughout the recording.
SimBA - Simple Behavioral Analysis
Open-source pipeline and GUI for developing supervised behavioral classifiers. Supports DLC, DeepPoseKit, SLEAP, MARS and others.
Bonsai
Visual programming language for synchronizing and processing heterogeneous data streams in real time
LabStreamingLayer (LSL)
Middleware framework for streaming and synchronizing multiple data sources (e.g. EEG, motion tracking, video, etc.) over a network.
Open Ephys
Open-source platform for electrophysiology hardware and software.
DeepLabCut
Open-source toolbox for markerless pose estimation of animals using deep learning, allowing high-accuracy tracking of user-defined body parts across species.
SLEAP (Social LEAP Estimates Animal Poses)
Open-source multi-animal pose tracking toolkit that uses deep learning to estimate positions of animal body landmarks, enabling tracking of multiple animals without markers.
DeepPoseKit
Python toolkit for animal pose estimation using deep learning, providing a user-friendly way to train networks for tracking body landmarks in videos.
JAABA (Janelia Automatic Animal Behavior Annotator)
Machine learning-based software for analyzing and classifying animal behaviors from video data; allows users to train classifiers to recognize specific behaviors (originally developed for fruit fly studies).
MARS (Mouse Action Recognition System)
Deep learning pipeline for automated pose estimation and social behavior classification in mice. MARS tracks multiple mice and identifies behaviors like attack, mounting, and investigation in pairwise interactions.
BENTO (Behavior Ensemble and Neural Trajectory Observatory)
Matlab-based interface for synchronizing and visualizing multimodal data streams (video, pose, neural recordings, audio, etc.), allowing researchers to link neural activity with behavioral events.
Motion Sequencing (MoSeq)
Unsupervised behavior analysis method that decomposes 3D animal motion into sequences of repeatable modules or 'syllables'. Provides tools for depth-video capture and analysis to characterize spontaneous behaviors in mice.
OpenPose
Real-time multi-person 2D pose estimation library. OpenPose detects human body, hand, and facial keypoints in video, widely used for analyzing human posture and movements in natural interactions.
OpenFace
Open-source tool for facial behavior analysis. It can track facial landmarks, head pose, eye gaze, and facial action units, facilitating analysis of expressions and attention in human social interactions.
DeepSqueak
Deep learning-based software for detecting and analyzing ultrasonic vocalizations. Commonly used to automatically identify rodent calls in audio recordings, aiding studies of communication and affective state.
Anipose
Open-source toolkit for 3D pose estimation from multiple camera views.
Models
Pre-trained or reference AI/ML models.
DeepLabCut Model Zoo (Mackenzie Mathis, Hugging Face)
A collection of DeepLabCut pre-trained models for a wide array of animals, maintained by Mackenzie Mathis. Includes models for quadrupeds, mice, birds, humans, macaques, primates, horses, and more. See the Hugging Face profile for the full list and details.
DeepLabCut Organization (Hugging Face)
The official DeepLabCut organization on Hugging Face, hosting models and demos for animal pose estimation. Supports a wide array of species and provides links to the Model Zoo and related resources.
CEBRA: Consistent EmBeddings of high-dimensional Recordings using Auxiliary variables
Machine-learning method that can be used to compress time series in a way that reveals otherwise hidden structures in the variability of the data
Datasets
Collections of behavioral, neural, or multimodal data.
SuperAnimal-Quadruped-80K
Dataset aggregating quadruped animal images.
SuperAnimal-TopViewMouse-5K
Dataset aggregating top-view mouse images.
Benchmarks
Evaluation suites or tasks used to compare models or methods.
DeepLabCut Benchmarks
Various benchmarks for animal pose estimation including TRI-MOUSE, PARENTING-MOUSE, MARMOSETS, FISH, HORSE-30.
Animal Pose Estimation Leaderboard
Comparative leaderboard for animal pose estimation methods.
DeepLabCut Leaderboard
Comparative leaderboard for DLC benchmarks.
Papers & Protocols
Academic publications, protocols, or documentation for methods.
Using DeepLabCut for 3D Markerless Pose Estimation
Guidance on DLC for 3D pose estimation.
DeepLabCut: Markerless Pose Estimation
Foundational paper on DLC technology.
DeeperCut: Multi-person Pose Estimation
High-performance multi-person pose estimation.
BRAIN Initiative Data Sharing Ecosystem (2024)
Overview and analysis of the NIH BRAIN Initiative's data archive ecosystem, comparing seven major neurodata repositories and discussing challenges in making neuroscience data more FAIR (Findable, Accessible, Interoperable, Reusable).
Platforms & Archives
Platforms or infrastructures that host, organize, or distribute data/tools.
OpenBehavior
Community-driven repository and blog highlighting open-source neuroscience tools and behavioral research methods. OpenBehavior aggregates resources for tracking, analysis, and hardware across species.
OpenNeuro
Free and open platform for sharing neuroscience data (primarily human brain imaging and electrophysiology) in BIDS format.
DANDI Archive
Data archive for neurophysiology and neuroimaging data, supporting the Neurodata Without Borders (NWB) format. DANDI enables sharing and reuse of large multimodal neuroscience datasets, aligning with FAIR data principles.
NIMH Data Archive (NDA)
Central repository for human subjects data (e.g., ABCD study) including behavioral, clinical, and neuroimaging data. NDA operates under controlled access with data use agreements, providing governance for sharing sensitive human data.
Standards & Ontologies
Definitions, formats, and ontologies to structure and share data consistently.
Brain Imaging Data Structure (BIDS)
A standard for organizing and describing neuroscience data to facilitate sharing and analysis.
Neurodata Without Borders (NWB)
Open standard format for neurophysiology data (neural recordings, behavioral events, etc.) to enhance sharing and integration.
Neuro Behavior Ontology (NBO)
Ontology of human and animal behaviors and behavioral phenotypes.
Hierarchical Event Descriptors (HED)
Standardized tagging system for events and metadata in experiments. HED tags provide a consistent way to label events (e.g., stimuli, actions) in neurobehavioral datasets, improving interoperability and analysis.
NIH BRAIN Initiative Neuroethics Guiding Principles
Eight principles outlining ethical considerations for brain research, emphasizing safety, autonomy, privacy of neural data, and public engagement.
Open Brain Consent
Initiative providing standardized consent language and tools to facilitate sharing of human brain data.
Contribution
To add resources or update existing entries, please reach out to the BBQS team.
Future Plans
We are working on integrating a chatbot that will connect to a database of JSON-organized resources, making it even easier to find the information you need. We are also working on an ingestion tool that will make it easier to add and update entries! Stay tuned for updates.