Synthetic Handwritten Digit Images, 10,000 Samples on Black Background

A synthetic dataset of 10,000 images in .jpg format, each with a resolution of 64 by 64 pixels. It contains 1,000 images for each digit from 0 to 9, all presented on a black background. The dataset was created by author MishaGGG and was last updated on the platform in June 2026.

Use Cases

• Train a digit classifier based on the 10,000 labeled images of digits 0-9.
• Benchmark model performance on synthetic data based on the uniform 64x64 pixel resolution and black background.
• Test data augmentation techniques based on the consistent image format and class balance.

Strengths

• Contains 10,000 total images, providing a substantial sample size for model training.
• Offers a balanced distribution with 1,000 images per digit class from 0 to 9.
• All images share a consistent 64x64 pixel resolution and black background, simplifying preprocessing.

Limitations

• The dataset is synthetic, which may limit its realism compared to natural handwritten digit data.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

huggingface

Collection Method

Synthetically generated, as stated in the description.

Freshness

Last updated 2026-06-01 08:10:47; freshness should be verified.