Training Hyperparameters and Accuracy for Rice Leaf Disease Detection Models

1914 image samples of rice leaves were processed to evaluate five deep transfer learning architectures. The dataset, created by Ummer Shakeel and last updated in May 2026, contains model training hyperparameters and configuration details, with results including test accuracies up to 98.43% for InceptionV3. It is shared under a CC-BY-4.0 license on figshare.

Use Cases

• Benchmarking deep learning model performance based on reported training, validation, and test accuracies.
• Comparing hyperparameter configurations for transfer learning architectures like InceptionV3 and MobileNetV2.
• Reproducing experimental results for rice leaf disease classification as described in the associated study.
• Analyzing the trade-off between model complexity and accuracy for lightweight models like MobileNetV2.

Strengths

• Includes specific performance metrics for five models, such as InceptionV3's 98.43% test accuracy.
• Based on a balanced dataset of 1914 image samples processed with TensorFlow and GPU acceleration.
• Shared under a permissive CC-BY-4.0 license, facilitating reuse and reproduction.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for large-scale analysis.
• The dataset is very small at 5.5 KB, indicating it likely contains summary parameters, not the raw image data.

Provenance

Source

figshare

Collection Method

Systematic processing of a public rice leaf disease image dataset using Python with TensorFlow.

Time Range

The study involved training models over 30 epochs; specific data collection date is not provided.

Freshness

Last updated 2026-05-07 17:40:55; freshness should be verified.

Geography

The research context is Pakistan, a major rice producer, but the geographic scope of the underlying image data is not specified.