Cataract Eye Images for Mobile Screening, 13K+ Images

13,000+ eye images labeled for cataract detection, organized into three classes. The dataset was designed for mobile screening applications and includes benchmark results from five convolutional neural networks. It is hosted on Kaggle.

Use Cases

• Train a cataract detection classifier based on the three-class image labels.
• Benchmark convolutional neural network performance on a medical image dataset.
• Develop lightweight vision models suitable for mobile deployment based on the dataset's mobile screening focus.

Strengths

• Contains over 13,000 labeled images, providing a substantial volume for model training.
• Includes benchmark results from five different CNN architectures, offering a performance baseline.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect geographic, temporal, or source bias inherent to Kaggle.

Provenance

Source

Kaggle