<p>Performance Statistics Across Four Random Seeds.</p>

Pakistan is the fourth-largest rice producer and the fifth-largest exporter worldwide. This 5.5 KB dataset by Ummer Shakeel, last updated in May 2026, contains performance statistics from a comparative evaluation of five deep transfer learning architectures for classifying rice leaf diseases. The models, including InceptionV3 and MobileNetV2, were trained on a balanced dataset of 1914 image samples.

Use Cases

• Benchmarking deep transfer learning models based on reported training, validation, and test accuracy metrics.
• Comparing lightweight versus dense neural network architectures for agricultural image classification based on the described models.
• Analyzing the impact of post-hoc visualization techniques like GrabCut segmentation on model interpretability as mentioned in the description.

Strengths

• Performance metrics for five distinct deep learning architectures (InceptionV3, DenseNet201, ResNet152V2, EfficientNetV2L, MobileNetV2) are provided.
• Models were trained over 30 epochs on a balanced dataset of 1914 image samples.
• Results include training, validation, and test accuracy percentages, such as InceptionV3 achieving 98.43% test accuracy.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is very small at 5.5 KB, indicating limited scope.

Provenance

Source

figshare

Collection Method

Results from a systematic comparative evaluation of deep learning models trained on a publicly available rice leaf disease image dataset.

Time Range

The study likely reflects the time of the computational experiments, but the exact temporal coverage of the underlying image data is not specified.

Freshness

Last updated 2026-05 07 17:41:11; freshness should be verified.

Geography

The research context is Pakistan's rice cultivation, but the geographic origin of the specific image samples used is not stated.