Hand Keypoints Dataset

Introdução

The hand-keypoints dataset contains 26,768 images of hands annotated with keypoints, making it suitable for training models like Ultralytics YOLO for pose estimation tasks. The annotations were generated using the Google MediaPipe library, ensuring high accuracy and consistency, and the dataset is compatible Ultralytics YOLOv8 formats.

Hand Landmarks

KeyPoints

The dataset includes keypoints for hand detection. The keypoints are annotated as follows:

1. Wrist
2. Thumb (4 points)
3. Index finger (4 points)
4. Middle finger (4 points)
5. Ring finger (4 points)
6. Little finger (4 points)

Each hand has a total of 21 keypoints.

Características principais

• Large Dataset: 26,768 images with hand keypoint annotations.
• YOLOv8 Compatibility: Ready for use with YOLOv8 models.
• 21 Keypoints: Detailed hand pose representation.

Estrutura do conjunto de dados

The hand keypoint dataset is split into two subsets:

1. Train: This subset contains 18,776 images from the hand keypoints dataset, annotated for training pose estimation models.
2. Val: This subset contains 7992 images that can be used for validation purposes during model training.

Aplicações

Hand keypoints can be used for gesture recognition, AR/VR controls, robotic manipulation, and hand movement analysis in healthcare. They can be also applied in animation for motion capture and biometric authentication systems for security.

Conjunto de dados YAML

A YAML (Yet Another Markup Language) file is used to define the dataset configuration. It contains information about the dataset's paths, classes, and other relevant information. In the case of the Hand Keypoints dataset, the hand-keypoints.yaml é mantido em https://github.com/ultralytics/ultralytics/blob/main/ultralytics/cfg/datasets/hand-keypoints.yaml.

# Ultralytics YOLO 🚀, AGPL-3.0 license
# Hand Keypoints dataset by Ultralytics
# Documentation: https://docs.ultralytics.com/datasets/pose/hand-keypoints/
# Example usage: yolo train data=hand-keypoints.yaml
# parent
# ├── ultralytics
# └── datasets
#     └── hand-keypoints  ← downloads here (369 MB)

# Train/val/test sets as 1) dir: path/to/imgs, 2) file: path/to/imgs.txt, or 3) list: [path/to/imgs1, path/to/imgs2, ..]
path: ../datasets/hand-keypoints # dataset root dir
train: train # train images (relative to 'path') 210 images
val: val # val images (relative to 'path') 53 images

# Keypoints
kpt_shape: [21, 3] # number of keypoints, number of dims (2 for x,y or 3 for x,y,visible)
flip_idx:
  [0, 1, 2, 4, 3, 10, 11, 12, 13, 14, 5, 6, 7, 8, 9, 15, 16, 17, 18, 19, 20]

# Classes
names:
  0: hand

# Download script/URL (optional)
download: https://github.com/ultralytics/assets/releases/download/v0.0.0/hand-keypoints.zip

Utilização

To train a YOLOv8n-pose model on the Hand Keypoints dataset for 100 epochs with an image size of 640, you can use the following code snippets. For a comprehensive list of available arguments, refer to the model Training page.

from ultralytics import YOLO

# Load a model
model = YOLO("yolov8n-pose.pt")  # load a pretrained model (recommended for training)

# Train the model
results = model.train(data="hand-keypoints.yaml", epochs=100, imgsz=640)

# Start training from a pretrained *.pt model
yolo pose train data=hand-keypoints.yaml model=yolov8n-pose.pt epochs=100 imgsz=640

Exemplos de imagens e anotações

The Hand keypoints dataset contains a diverse set of images with human hands annotated with keypoints. Here are some examples of images from the dataset, along with their corresponding annotations:

• Imagem em mosaico: Esta imagem demonstra um lote de treino composto por imagens do conjunto de dados em mosaico. O mosaico é uma técnica utilizada durante o treino que combina várias imagens numa única imagem para aumentar a variedade de objectos e cenas em cada lote de treino. Isto ajuda a melhorar a capacidade do modelo de generalizar para diferentes tamanhos de objectos, proporções e contextos.

The example showcases the variety and complexity of the images in the Hand Keypoints dataset and the benefits of using mosaicing during the training process.

Citações e agradecimentos

If you use the hand-keypoints dataset in your research or development work, please acknowledge the following sources:

We would also like to acknowledge the creator of this dataset, Rion Dsilva, for his great contribution to Vision AI research.

📅 Created 0 days ago ✏️ Updated 0 days ago

Comentários