Skip to content

Dataset · ntsrigaud

Balanced hand gesture subset from HaGRID

ntsrigaud/hagrid-subset

Cleaned subset of the famous HaGRID dataset with 19,200 photos for 24 hand gesture recognition and detection.

What's inside

The dataset contains 19,200 JPEG images with an average size of about 302 KB. The data is strictly balanced: 800 examples for each of the 24 gesture classes. The set is already split into training (14,592 photos), validation (1,728), and test (2,880) splits.

What you can do with it

  • Training gesture detectors and classifiers for contactless control systems
  • Rapid prototyping and hypothesis testing on a balanced version of HaGRID
  • Fine-tuning lightweight mobile computer vision models for gesture recognition
Tasks
Image classification, Object detection
Size
10K–100K rows
Downloads this month
27.6K
Likes
1

How to load it

from datasets import load_dataset

ds = load_dataset("ntsrigaud/hagrid-subset")