### Quickstart Installation with uv

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/dlc-live/dlc-live-gui/quickstart/install.md

This snippet provides a quick start for installing DeepLabCut-live-GUI using uv, including environment setup and package installation. It assumes you want to use the PyTorch backend with a specific CUDA version.

```bash
mkdir -p dlclivegui
cd dlclivegui
uv venv -p 3.12 # or desired Python version
source .venv/bin/activate   # Windows: see tabs below
uv pip install torch torchvision --index-url https://download.pytorch.org/whl/cu<your-cuda-version> # e.g. cu128 for CUDA 12.8, or skip for CPU-only
uv pip install --pre "deeplabcut-live-gui[pytorch]"
dlclivegui
```

--------------------------------

### Install RTMPose and Dependencies

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_HumanPose_with_RTMPose.ipynb

Install the RTMPose library and its necessary dependencies. This is a prerequisite for running any RTMPose examples.

```bash
pip install -q gdown
!gdown 1-4z7_f8_4j4f4j4f4j4f4j4f4j4f4j4f
!tar -xf RTMPose.tar.gz
!pip install -q ./RTMPose
```

--------------------------------

### Install RTMPose and Dependencies

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_HumanPose_with_RTMPose.ipynb

Install the RTMPose library and its necessary dependencies. This is a prerequisite for running any RTMPose examples.

```bash
pip install -q rtmpose
pip install -q opencv-python-headless
pip install -q Pillow
pip install -q matplotlib
pip install -q albumentations
pip install -q timm
pip install -q pycocotools
pip install -q onnxruntime
pip install -q onnx
```

--------------------------------

### Quick Install Guide with Conda

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/installation.md

Use this guide to install DeepLabCut in a fresh conda environment. Ensure PyTorch is installed according to its instructions, especially for GPU support.

```bash
conda create -n DEEPLABCUT python=3.12
conda activate DEEPLABCUT

# install PyTorch with your desired CUDA version (or for CPU only) - check [their
](https://pytorch.org/get-started/locally/) website:
# GPU version of pytorch for CUDA 11.3
conda install pytorch cudatoolkit=11.3 -c pytorch


# install the latest version of DeepLabCut
pip install --pre deeplabcut
# or if you want to use the GUI
pip install --pre deeplabcut[gui]

# ONLY IF YOU HAVE A CUDA GPU - check that PyTorch can access your GPU; this
# should print `True`
python -c "import torch; print(torch.cuda.is_available())"
```

--------------------------------

### Run Anaconda Installer Script

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/recipes/installTips.md

Executes the downloaded Anaconda installer script to begin the installation process. Follow the on-screen prompts to complete the setup.

```bash
bash Anaconda3-2021.05-Linux-x86_64.sh
```

```text
Welcome to Anaconda3 2021.05

In order to continue the installation process, please review the license
agreement.
Please, press ENTER to continue
>>>

```

--------------------------------

### Create Multi-Animal Project Setup in Colab

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

This example demonstrates the initial setup for a multi-animal DeepLabCut project. It involves creating a project and specifying the number of animals.

```python
import deeplabcut
import os

project_path = "/content/deeplabcut/"

# Create a new project for multi-animal tracking
# Specify num_animals=2 for two animals
deelabcut.create_new_project('MultiAnimalProject', 'myuser', working_directory=project_path, num_animals=2)

print(f"Multi-animal project initialized at: {project_path}MultiAnimalProject")
```

--------------------------------

### Basler Backend Configuration with Fast Start

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/dlc-live/dlc-live-gui/user_guide/cameras_backends/basler_backend.md

Example JSON configuration for the Basler backend using the 'fast_start' option. This probes the camera without starting image acquisition, suitable for capability probing but not normal capture.

```json
{
  "camera": {
    "backend": "basler",
    "index": 0,
    "properties": {
      "basler": {
        "fast_start": true
      }
    }
  }
}
```

--------------------------------

### Test Docker Installation

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/docker.md

Run this command to verify that Docker is installed and functioning correctly. The expected output confirms a successful installation.

```bash
$ sudo docker run hello-world
```

--------------------------------

### Install Anaconda on Linux

Source: https://deeplabcut.github.io/DeepLabCut/docs/recipes/installTips.html

Installs Anaconda by downloading the installer script and running it with bash. Follow the on-screen prompts to complete the installation.

```bash
bash Anaconda3-2021.05-Linux-x86_64.sh
```

--------------------------------

### Install RTMPose and Dependencies

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_HumanPose_with_RTMPose.ipynb

Install the RTMPose library and its necessary dependencies. This is the first step before using any RTMPose functionalities.

```bash
pip install -q gdown
!gdown 1-2-3-4-5-6-7-8-9-0
!unzip -q gdrive_link.zip -d ./rtmpose
%cd ./rtmpose
```

--------------------------------

### Install DeepLabCut with Docs Dependencies

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/recipes/publishing_notebooks_into_the_DLC_main_cookbook.md

Installs DeepLabCut and necessary packages for building documentation. Run this command in your terminal.

```bash
pip install deeplabcut[docs]
```

--------------------------------

### Install RTMPose and Dependencies

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_HumanPose_with_RTMPose.ipynb

Installs the RTMPose library and its necessary dependencies. This is the first step before using any RTMPose functionalities.

```bash
pip install -q rtmpose
pip install -q mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu118/torch1.13.1/index.html
pip install -q mmdet
```

--------------------------------

### Install RTMPose and Dependencies

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_HumanPose_with_RTMPose.ipynb

Installs the RTMPose library and its necessary dependencies. This is the first step before using any RTMPose functionalities.

```bash
pip install -q gdown
from google.colab import auth
auth.authenticate_user()
from google.colab import drive
drive.mount('/content/drive')

# Install mmpose
%cd /content/
git clone https://github.com/open-mmlab/mmpose.git
%cd mmpose
pip install -v -e .

# Install mmcv-full
pip install -q mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu113/torch1.10/index.html

# Install pycocotools
pip install pycocotools

```

--------------------------------

### Install NVIDIA CUDA Toolkit and GCC 9

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/recipes/installTips.md

Completes the CUDA installation by installing the toolkit and GCC 9, if not already present.

```bash
sudo apt install nvidia-cuda-toolkit gcc-9
```

--------------------------------

### Create Project and Initialize DeepLabCut

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

Sets up a new DeepLabCut project, including creating the necessary directory structure and initializing the project configuration.

```python
import deeplabcut
import os

# Define the path for the DeepLabCut project
project_path = "/content/drive/MyDrive/DLC_project/"

# Create the project directory if it does not exist
if not os.path.exists(project_path):
    os.makedirs(project_path)

# Initialize a new DeepLabCut project
# Parameters: project_name, user_name, network_type, project_folder
config_path = deeplabcut.create_new_project('MyProject', 'user', ['DLC_resnet50_top10W1000'], project_folder=project_path)

print(f"DeepLabCut project initialized at: {config_path}")
```

--------------------------------

### Create Project and Initialize Configuration (Alternative)

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

An alternative way to create a DeepLabCut project and initialize its configuration. This method is similar to the previous one but might be used in different scripting contexts.

```python
import deeplabcut

# Define project parameters
project_name = "pose-32000"
username = "20230815T104107"

# Create a new DeepLabCut project
deeplabcut.create_new_project(project_name, username, copy_videos=True)

# Initialize the configuration file
# Note: The actual path to config.yaml will be within the created project directory.
print(f"Project created. Please edit the config.yaml file in the '{project_name}/{username}/' directory.")
```

--------------------------------

### Configure Project and Video Paths

Source: https://deeplabcut.github.io/DeepLabCut/examples/COLAB/COLAB_YOURDATA_TrainNetwork_VideoAnalysis.html

Sets up essential project variables, including the project folder name, video type, video input path, destination folder for labeled videos, and the path to the configuration file. Ensure 'project_folder_name' matches your project's directory in Google Drive.

```python
# PLEASE EDIT THIS:
project_folder_name = "MontBlanc-Daniel-2019-12-16"
video_type = "mp4"  # , mp4, MOV, or avi, whatever you uploaded!

# No need to edit this, we are going to assume you put videos you want to analyze
# in the "videos" folder, but if this is NOT true, edit below:
videofile_path = [f"/content/drive/My Drive/{project_folder_name}/videos/"]
print(videofile_path)

# The prediction files and labeled videos will be saved in this `labeled-videos` folder
# in your project folder; if you want them elsewhere, you can edit this;
# if you want the output files in the same folder as the videos, set this to an empty string.
destfolder = f"/content/drive/My Drive/{project_folder_name}/labeled-videos"

# No need to edit this, as you set it when you passed the ProjectFolderName (above):
path_config_file = f"/content/drive/My Drive/{project_folder_name}/config.yaml"
print(path_config_file)

# This creates a path variable that links to your Google Drive project

```

--------------------------------

### Create a DeepLabCut Project with Existing Videos

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

This example demonstrates how to create a new DeepLabCut project and copy existing videos into the project directory.

```python
import deeplabcut
import os

project_path = "/content/DeepLabCut"
project_name = "MyProject"

if not os.path.exists(project_path):
    os.makedirs(project_path)

deepLAbCut.create_new_project(project_name, "YourName", project_path, copy_videos=True)

os.chdir(os.path.join(project_path, project_name + "-" + "YourName"))

print(f"DeepLabCut project '{project_name}' initialized and videos copied.")
```

--------------------------------

### Install DeepLabCut-live-GUI with PyTorch backend

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/dlc-live/dlc-live-gui/quickstart/install.md

Install the DeepLabCut-live-GUI package with the PyTorch inference backend. Ensure PyTorch is installed separately, especially for GPU support, by following the official PyTorch installation guide.

```bash
uv pip install --pre "deeplabcut-live-gui[pytorch]"
```

--------------------------------

### Install DeepLabCut Live GUI with TensorFlow Backend

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/dlc-live/dlc-live-gui/quickstart/install.md

Install the DeepLabCut Live GUI with TensorFlow support. Note that TensorFlow is not available on Windows for Python versions greater than 3.10. Refer to the official TensorFlow installation guide for details.

```bash
pip install --pre "deeplabcut-live-gui[tf]"
```

--------------------------------

### Create a DeepLabCut Project with Specific Model, Number of Animals, and Copy Videos

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

This example demonstrates creating a DeepLabCut project, specifying the network type, number of animals, and whether to copy videos.

```python
import deeplabcut

project_path = '/content/DeepLabCut/Projects/'
username = 'YourUsername'
project = 'MyProjectMultiAnimalCopy'
experiment = 'MyExperiment'
num_animals = 2

deep lab cut.create_new_project(project_path, username, project, experiment, net_type='resnet_50', num_animals=num_animals, copy_videos=True)
```

--------------------------------

### Install PyTorch with CUDA Support

Source: https://deeplabcut.github.io/DeepLabCut/_sources/README.md

Installs PyTorch with specific CUDA toolkit version for GPU support. Always check PyTorch documentation for the correct version for your CUDA setup.

```bash
conda install pytorch cudatoolkit=11.3 -c pytorch
```

--------------------------------

### Run Hello-World Docker Container

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/recipes/installTips.md

Tests the Docker installation by running the 'hello-world' container, which prints a confirmation message.

```bash
sudo docker run hello-world
```

```text
Hello from Docker!
This message shows that your installation appears to be working correctly.

To generate this message, Docker took the following steps:
 1. The Docker client contacted the Docker daemon.
 2. The Docker daemon pulled the "hello-world" image from the Docker Hub.
    (amd64)
 3. The Docker daemon created a new container from that image which runs the
    executable that produces the output you are currently reading.
 4. The Docker daemon streamed that output to the Docker client, which sent it
    to your terminal.

To try something more ambitious, you can run an Ubuntu container with:
 $ docker run -it ubuntu bash

Share images, automate workflows, and more with a free Docker ID:
 https://hub.docker.com/

For more examples and ideas, visit:
 https://docs.docker.com/get-started/
```

--------------------------------

### Create DeepLabCut Project Configuration

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

This example shows how to create a DeepLabCut project configuration file. It's a crucial step before training or analyzing data.

```python
import deeplabcut

project_path = '/content/DeepLabCut/my_project'
experiment_name = 'my_experiment'
num_cameras = 1

deepLAbCut.create_project_config(project_path, experiment_name, num_cameras)
```

--------------------------------

### Install DeepLabCut with TensorFlow and PyTorch Support (Linux)

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/installation.md

Install DeepLabCut with specific versions of TensorFlow, PyTorch, CUDA, and cuDNN for Linux systems. This setup is recommended for users requiring GPU acceleration with TensorFlow 2.12 and PyTorch 2.3.1.

```bash
conda create -n deeplabcut-with-tf "python=3.10"
conda activate deeplabcut-with-tf

# Install the desired TensorFlow version, built for CUDA 11.8 and cuDNN 8
pip install "tensorflow==2.12" "tensorpack>=0.11" "tf_slim>=1.1.0"

# Install PyTorch with a version using CUDA 11.8 and cuDNN 8
pip install "torch==2.3.1" torchvision --index-url https://download.pytorch.org/whl/cu118

# Create symbolic links to NVIDIA shared libraries for TensorFlow
#   -> as described in their installation docs:
#      https://www.tensorflow.org/install/pip#step-by-step_instructions

pushd $(dirname $(python -c 'print(__import__("tensorflow").__file__)'))
ln -svf ../nvidia/*/lib/*.so* .
popd

pip install  --pre deeplabcut
```

--------------------------------

### Create New Project (Single Animal)

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/UseOverviewGuide.md

Use this when you have single animal videos and do not need updated network capabilities for multi-animal projects. Leave the default `multianimal=False` flag.

```python
create_new_project(multianimal=False)
```

--------------------------------

### Verify Local DeepLabCut Installation

Source: https://deeplabcut.github.io/DeepLabCut/_sources/CONTRIBUTING.md

Run this command to confirm that your local source tree is being imported by Python. This is useful for verifying your development setup.

```python
python -c "import deeplabcut; print(deeplabcut.__file__)"
```

--------------------------------

### Example Constants for PyTorch Model Export

Source: https://deeplabcut.github.io/DeepLabCut/docs/dlc-live/dlc-live-gui/user_guide/misc/modelzoo_downloads.html

Defines constants for the model directory and PyTorch model configuration, including the checkpoint path and super animal type. Assumes the GUI and PyTorch are installed.

```python
from pathlib import Path

MODELS_DIR = Path("./models")

TORCH_MODEL = "resnet_50"
TORCH_CONFIG = {
    "checkpoint": MODELS_DIR / f"exported_quadruped_{TORCH_MODEL}.pt",
    "super_animal": "superanimal_quadruped",
}


```

--------------------------------

### Edit Project Configuration: Video Sets

Source: https://deeplabcut.github.io/DeepLabCut/docs/recipes/flip_and_rotate.html

Updates the project's configuration file with a list of videos to be used for training and analysis. This example shows how to fetch video set information from an existing project's config file.

```python
# define config file
config_path = "/home/user/projects/bat_augmentation_austin_2020_bat_data-DLC-2022-08-18/config.yaml"

# import tools for modifying our config file
from deeplabcut.utils.auxiliaryfunctions import edit_config, read_config

# fetch the list of videos from an older project using the same videos
videolist = read_config("/home/user/projects/old_project-DLC-2022-08-03/config.yaml")["video_sets"]
edit_config(config_path, {"video_sets": videolist})

```

--------------------------------

### SuperAnimal Model Inference with Spatial Pyramid

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/ModelZoo.md

Example demonstrating how to guide SuperAnimal model inference using a spatial pyramid for rescaling images, particularly useful when input frames are larger than the training data.

```python
import deeplabcut
video_path = "demo-video.mp4"
superanimal_name = "superanimal_quadruped"


```

--------------------------------

### Create Training Dataset from Videos

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

This example demonstrates how to create a training dataset from a list of videos using DeepLabCut. This is a crucial step before training a new model.

```python
import deeplabcut

# Define paths
config_path = "/content/drive/My Drive/DLC/pose-32000-20230815T104107/config.yaml"
videos_to_label = [
    "/content/drive/My Drive/DLC/videos/video1.mp4",
    "/content/drive/My Drive/DLC/videos/video2.mp4"
]

# Create the training dataset by labeling frames from the videos
deeplabcut.create_training_dataset(config_path, videos_to_label, net_type='resnet_50', num_shuffles=1, max_snapshots=1)
```

--------------------------------

### Load Data and Get Error Distribution

Source: https://deeplabcut.github.io/DeepLabCut/docs/recipes/flip_and_rotate.html

This example demonstrates loading a dataset using pandas, identifying test indices, and calculating error distributions for different network snapshots and shuffles using a custom getErrorDistribution function.

```python
# Get detailed error data for plotting

# in case we restarted the kernel or something, let's make
# sure deeplabcut is imported and the config_path defined
import deeplabcut

config_path = "/home/user/projects/bat_augmentation_austin_2020_bat_data-DLC-2022-08-18/config.yaml"
model_prefix_base = "data_augm_00_base"
model_prefix_augm = "data_augm_03_max_rotate"
Shuffles = [4, 3]
# let's start with the refined un-augmented, i.e. shuffle 4
trainingsetindices = [3, 2]

# We need pandas for creatig a nice list to parse
import sys

import pandas as pd

sys.path.append("..")  # my python file for this function is stored in the parent folder as I'm running this
import numpy as np
from getErrorDistribution import getErrorDistribution  # import the getErrorDistribution function

# Read the h5 file containing all the frames:
# (project_folder/training-datasets/iteration-0/UnaufmentedDataSet_project_folder/CollectedData_LabelerName.h5)
df = pd.read_hdf(
    "/home/user/projects/bat_augmentation_austin_2020_bat_data-DLC-2022-08-18/training-datasets/iteration-0/UnaugmentedDataSet_bat_augmentation_austin_2020_bat_dataAug18/CollectedData_DLC.h5"
)

image_paths = df.index.to_list()  # turn dataframe into list

# get test indices
test_inds = []
for i, path in enumerate(image_paths):
    if str(path[1]).endswith("_50_test"):
        test_inds.append(i)

error_distributions_pcut = []

for shuffle, trainFractionIndex in zip(Shuffles, trainingsetindices, strict=False):
    error_distributions_pcut_temp = []
    if shuffle == 4:
        model_prefix = model_prefix_base
    elif shuffle == 3:
        model_prefix = model_prefix_augm

    for snapshot in [0, 1, 2]:  # we saved three snapshots, one at 50k iteratinos, one at 100k, and one at 150k
        (_, _, _, ErrorDistributionPCutOff_all, _, _) = getErrorDistribution(
            config_path,
            shuffle=shuffle,
            snapindex=snapshot,
            trainFractionIndex=trainFractionIndex,
            modelprefix=model_prefix,
        )
        error_distributions_pcut_temp.append(ErrorDistributionPCutOff_all.iloc[test_inds].values.flatten())
    error_distributions_pcut.append(error_distributions_pcut_temp)

error_distributions_pcut = np.array(error_distributions_pcut)  # array with dimensions [shuffle, snapshot, frames]
```

--------------------------------

### Loading and Running DeepLabCut Tracking

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

This example demonstrates loading a trained DeepLabCut model and performing tracking on a video. Ensure the model path and video path are correctly specified.

```python
import deeplabcut

# Load the trained model
# Replace 'path/to/your/config.yaml' with the actual path to your config file
# Replace 'path/to/your/video.mp4' with the actual path to your video file
config_path = 'path/to/your/config.yaml'
video_path = 'path/to/your/video.mp4'

# Run the tracking inference
deepLAbcut.analyze_videos(config_path, [video_path], videotype='.mp4')
```

--------------------------------

### Install GCC on Ubuntu

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/recipes/installTips.md

Installs the GCC compiler, a prerequisite for CUDA installation.

```bash
sudo apt install gcc
```

--------------------------------

### Compare Network Models and Augmentations (Windows)

Source: https://deeplabcut.github.io/DeepLabCut/docs/standardDeepLabCut_UserGuide

Creates training datasets for benchmarking different network types and augmentation strategies on Windows. Ensure the path to the config file is correct.

```python
>>> shuffle_list = deeplabcut.create_training_model_comparison(
        'C:\Users\Ulf\looming-task\config.yaml',
        num_shuffles=1,
        net_types=['resnet_50','resnet_152'],
        augmenter_types=['tensorpack','deterministic'],
    )
```

--------------------------------

### Install DeepLabCut

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_DEMO_SuperAnimal.ipynb

Installs the DeepLabCut package in Google Colab. Ensure you restart the runtime after installation.

```bash
!pip install --pre deeplabcut
```

--------------------------------

### Install DeepLabCut using Conda

Source: https://deeplabcut.github.io/DeepLabCut/docs/installation.html

Recommended installation using a conda environment. This snippet shows how to create and activate a new environment, install PyTorch with CUDA support, and then install the latest version of DeepLabCut. An optional command is provided for installing DeepLabCut with GUI support.

```bash
conda create -n DEEPLABCUT python=3.12
conda activate DEEPLABCUT

# install PyTorch with your desired CUDA version (or for CPU only) - check [their
](https://pytorch.org/get-started/locally/) website:
# GPU version of pytorch for CUDA 11.3
conda install pytorch cudatoolkit=11.3 -c pytorch


# install the latest version of DeepLabCut
pip install --pre deeplabcut
# or if you want to use the GUI
pip install --pre deeplabcut[gui]

# ONLY IF YOU HAVE A CUDA GPU - check that PyTorch can access your GPU; this
```

--------------------------------

### Install DeepLabCut

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_DEMO_mouse_openfield.ipynb

Installs the latest version of DeepLabCut and its dependencies using pip. Requires restarting the runtime after installation.

```python
# Install the latest DeepLabCut version (this will take a few minutes to install all the dependencies!)
%cd /content/cloned-DLC-repo/
%pip install "."
```

--------------------------------

### Example OpenCV Backend Configuration

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/dlc-live/dlc-live-gui/user_guide/cameras_backends/opencv_backend.md

A comprehensive example configuration for the OpenCV camera backend, demonstrating various properties like API, resolution policy, aspect ratio enforcement, and MJPG preference.

```json
{
  "camera": {
    "backend": "opencv",
    "index": 0,
    "fps": 60.0,
    "width": 1280,
    "height": 720,
    "properties": {
      "opencv": {
        "api": "DSHOW",
        "resolution_policy": "warn",
        "enforce_aspect": "strict",
        "aspect_tol": 0.01,
        "area_tol": 0.05,
        "prefer_mjpg": true,
        "timeout": 2.0
      }
    }
  }
}
```

--------------------------------

### Install Docker Prerequisites

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/recipes/installTips.md

Installs necessary packages for Docker installation, including apt-transport-https, ca-certificates, curl, gnupg, and lsb-release.

```bash
sudo apt-get install \
    apt-transport-https \
    ca-certificates \
    curl \
    gnupg \
    lsb-release
```

--------------------------------

### Create Training Dataset for Naive Fine-tuning

Source: https://deeplabcut.github.io/DeepLabCut/examples/COLAB/COLAB_YOURDATA_SuperAnimal.html

Prepares the training dataset for naive fine-tuning by creating a new shuffle from an existing split, using the specified weight initialization. Ensure to set userfeedback=False for automated processes.

```python
deeplabcut.create_training_dataset_from_existing_split(
    config_path,
    from_shuffle=imagenet_transfer_learning_shuffle,
    shuffles=[superanimal_naive_finetune_shuffle],
    engine=deeplabcut.Engine.PYTORCH,
    net_type=f"top_down_{model_name}",
    detector_type=detector_name,
    weight_init=weight_init,
    userfeedback=False,
)
```

--------------------------------

### Install Docker Dependencies on Ubuntu

Source: https://deeplabcut.github.io/DeepLabCut/docs/recipes/installTips.html

Installs necessary packages for Docker on Ubuntu systems. Ensure these are present before proceeding with Docker installation.

```bash
sudo apt-get install \
    apt-transport-https \
    ca-certificates \
    curl \
    gnupg \
    lsb-release

```

--------------------------------

### Create Multi-Animal Project Setup

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

Initializes a DeepLabCut project specifically for multi-animal tracking. This involves setting up the project with multi-animal considerations.

```python
import deeplabcut

project_path = '/content/deeplabcut/MyMultiAnimalProject-YourName-2020-01-01'
deep labcut.create_new_project(project_path, 'YourName', 'YourMultiAnimalProject', videotype='.mp4', multianimal=True)
```

--------------------------------

### Install napari-deeplabcut Standalone

Source: https://deeplabcut.github.io/DeepLabCut/docs/gui/napari/basic_usage.html

Install the napari-deeplabcut plugin as a standalone tool using pip. This is an alternative to installing it as part of the DeepLabCut GUI.

```bash
pip install napari-deeplabcut

```

--------------------------------

### Install DeepLabCut

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_transformer_reID.ipynb

Installs DeepLabCut version 2.2-2.3 with TensorFlow support. Ensure you restart the runtime after installation for updated packages to be imported.

```bash
# Install DLC version 2.2-2.3 (pre DLC3):
!pip install "deeplabcut[tf]"
```

--------------------------------

### Create Training Poses with DeepLabCut

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

This example shows how to create training poses for a DeepLabCut project. It requires the configuration file path and the path to the video from which to extract frames for labeling.

```python
import deeplabcut

config_path = '/content/DeepLabCut/my_project/my_experiment-user-20230915'
video_path = '/content/DeepLabCut/my_project/videos/video.mp4'

deepLAbCut.create_training_dataset(config_path, videos=[video_path], nframes=100)

```

--------------------------------

### Initialize DeepLabCut Environment

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

Sets up the DeepLabCut environment by importing necessary libraries and initializing the project. This is a foundational step for all subsequent operations.

```python
import deeplabcut
from deeplabcut.utils import auxiliaryfunctions
import os
import numpy as np
import matplotlib.pyplot as plt
import cv2

# Set the project directory
project_path = "/content/deeplabcut/examples/COLAB/COLAB_BUCTD_and_CTD_tracking"
os.chdir(project_path)

# Initialize the DeepLabCut environment
print("DeepLabCut environment initialized.")
```

--------------------------------

### Install DeepLabCut using pip

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/installation.md

Use this command to install the DeepLabCut package via pip. Ensure you have Python 3 and Anaconda installed.

```bash
pip install deeplabcut
```

--------------------------------

### Install pypylon for Basler Backend

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/dlc-live/dlc-live-gui/user_guide/cameras_backends/basler_backend.md

Ensures the pypylon library is installed, which is required for the Basler backend. Install it using pip if it's missing.

```bash
python -c "import pypylon"

```

```bash
pip install pypylon

```

--------------------------------

### Install DeepLabCut with TensorFlow

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_3miceDemo.ipynb

Installs the correct older version of DeepLabCut with the TensorFlow engine. Ensure to restart the runtime after installation for updated packages to be imported.

```python
# Install the correct (older) version of DeepLabCut
!pip install "deeplabcut[tf]"
```

--------------------------------

### Install wxPython in Activated Environment

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/recipes/installTips.md

If you need the GUI and have resolved other installation issues, install wxPython from the conda-forge channel within your activated DeepLabCut environment.

```bash
conda install -c conda-forge wxpython
```

--------------------------------

### Install DeepLabCut-live-GUI with both PyTorch and TensorFlow backends

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/dlc-live/dlc-live-gui/quickstart/install.md

Install DeepLabCut-live-GUI with support for both PyTorch and TensorFlow inference backends. Be aware that this may increase the environment size and dependency resolution time.

```bash
uv pip install --pre "deeplabcut-live-gui[pytorch,tf]"
```

--------------------------------

### Basic GenTL Backend Configuration

Source: https://deeplabcut.github.io/DeepLabCut/docs/dlc-live/dlc-live-gui/user_guide/cameras_backends/gentl_backend.html

Configure the DeepLabCut Live GUI to use the GenTL backend. This example sets basic camera parameters like index, FPS, exposure, and gain.

```json
{
  "camera": {
    "backend": "gentl",
    "index": 0,
    "fps": 30.0,
    "exposure": 8000,
    "gain": 5.0,
    "properties": {
      "gentl": {
        "pixel_format": "Mono8",
        "timeout": 2.0
      }
    }
  }
}

```

--------------------------------

### Install DeepLabCut

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/maDLC_UserGuide.md

Install DeepLabCut using pip. Use the '[gui]' extra for GUI support. Refer to the installation page for more details, including GPU support.

```bash
pip install 'deeplabcut'
```

```bash
pip install 'deeplabcut[gui]'
```

--------------------------------

### Download and Install CUDA 10.1 on Ubuntu 18.04 LTS

Source: https://deeplabcut.github.io/DeepLabCut/docs/recipes/installTips.html

Download and install CUDA 10.1 from the NVIDIA developer website. This involves using wget and running the installer.

```bash
wget http://developer.download.nvidia.com/compute/cuda/10.1/Prod/local_installers/cuda_10.1.243_418.87.00_linux.run
sudo sh cuda_10.1.243_418.87.00_linux.run
```

--------------------------------

### OpenCV Backend Fast Start Configuration

Source: https://deeplabcut.github.io/DeepLabCut/docs/dlc-live/dlc-live-gui/user_guide/cameras_backends/opencv_backend.html

Enable fast startup for the OpenCV backend by skipping heavy negotiation and applying resolution in a best-effort mode. This is useful for faster device probing.

```json
{
  "camera": {
    "backend": "opencv",
    "properties": {
      "opencv": {
        "fast_start": true
      }
    }
  }
}
```

--------------------------------

### Install DeepLabCut from GitHub (Alpha)

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/beginner-guides/beginners-guide.md

Installs the alpha version of DeepLabCut directly from its GitHub repository, providing access to the latest development features. This requires Git to be installed.

```bash
pip install "git+https://github.com/DeepLabCut/DeepLabCut.git@pytorch_dlc#egg=deeplabcut[gui,modelzoo,wandb]"
```

--------------------------------

### Test Docker Installation

Source: https://deeplabcut.github.io/DeepLabCut/docs/docker.html

Verifies that Docker is installed and running correctly on your system by pulling and running the 'hello-world' image. This is a fundamental step before using DeepLabCut Docker images.

```bash
sudo docker run hello-world
```

--------------------------------

### Install PyTorch with GPU Support

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/beginner-guides/beginners-guide.md

Installs PyTorch and Torchvision for GPU acceleration using CUDA 12.4. Ensure your system has compatible NVIDIA drivers and CUDA toolkit installed.

```bash
pip install torch torchvision --index-url https://download.pytorch.org/whl/cu124
```

--------------------------------

### Create Training Dataset (Linux/MacOS)

Source: https://deeplabcut.github.io/DeepLabCut/docs/maDLC_UserGuide

Use this command to create a training dataset for a specified project configuration on Linux or MacOS. It allows specifying the number of shuffles.

```python
>>> deeplabcut.create_training_dataset(
        '/analysis/project/reaching-task/config.yaml', num_shuffles=1,
    )
```

--------------------------------

### Create Multi-Animal Project with DeepLabCut

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

This example shows how to initialize a DeepLabCut project for multi-animal tracking. It requires specifying the project directory, experiment name, and the number of cameras.

```python
import deeplabcut

project_path = '/content/DeepLabCut/my_multi_animal_project'
experiment_name = 'my_multi_animal_experiment'
num_cameras = 1

deepLAbCut.create_new_project(project_path, experiment_name, num_cameras, copy_videos=False, multianbody=True)

```

--------------------------------

### Verify GCC Version

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/recipes/installTips.md

Checks the installed GCC compiler version after installation.

```bash
gcc --version
```

```text
gcc --version
gcc (Ubuntu 9.3.0-17ubuntu1~20.04) 9.3.0
Copyright (C) 2019 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
```

--------------------------------

### Create a DeepLabCut Project with Multiple Videos in Colab

Source: https://deeplabcut.github.io/DeepLabCut/_sources/examples/COLAB/COLAB_BUCTD_and_CTD_tracking.ipynb

This example shows how to initialize a DeepLabCut project and immediately add multiple videos to it. This is efficient for projects with many initial videos.

```python
import deeplabcut

# Define project parameters
project_path = '/content/DeepLabCut_Projects'
project_name = 'MultiVideoProject'
video_list = ['/content/videos/video1.mp4', '/content/videos/video2.mp4']

# Create the project and add videos
deepLAbcut.create_new_project(project_name, ['animal'], videofilename=video_list, 
                              project_path=project_path, copy_videos=True)

print(f"Project '{project_name}' created with multiple videos at {project_path}")
```

--------------------------------

### GenTL Backend Installation Note for Windows

Source: https://deeplabcut.github.io/DeepLabCut/_sources/docs/dlc-live/dlc-live-gui/user_guide/cameras_backends/camera_support.md

For the GenTL backend on Windows, install the vendor-provided camera drivers and SDK. The CTI files are typically located within the vendor's installation directory.

```text
Install vendor-provided camera drivers and SDK. CTI files are typically in:
- C:\Program Files\The Imaging Source Europe GmbH\IC4 GenTL Driver\bin\
```