PlantTom: Balanced Dataset of 7,836 Tomato Leaf Disease Images

7,836 images of tomato leaf diseases across 8 distinct classes, compiled from various public sources to address class imbalance. The dataset was created by Vaskor Mostafa and last updated on 2026-05-22. It was developed to support the LeafDet object detection model for smart agriculture applications.

Use Cases

• Training object detection models for tomato leaf disease identification based on the 8 distinct disease classes.
• Benchmarking model performance on a balanced dataset to mitigate class imbalance issues mentioned in the description.
• Developing interpretable AI frameworks for agriculture using visualization techniques like Eigen-CAM referenced in the paper.
• Comparing the efficacy of different YOLO architectures and attention modules for plant pathology tasks.

Strengths

• 7,836 images provide a substantial base for model training.
• 8 distinct, balanced classes address a common limitation in public plant disease datasets.
• Dataset is explicitly designed to support reproducible and less biased model evaluation.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for specific tasks.
• The 5.5 KB file size suggests the XLS file contains metadata or annotations, not the actual images, requiring separate image sourcing.

Provenance

Source

Combined from various public sources by author Vaskor Mostafa.

Collection Method

Curated and revised to create a balanced dataset named PlantTom.

Time Range

null

Freshness

Last updated 2026-05-22 18:00:31; freshness should be verified.

Geography

null