PlantaNetLite: Plant Disease Classification Model Training Results

PlantaNetLite is a training result summary for a plant disease classification framework evaluated on a curated dataset of approximately 45,000 images across 51 disease and healthy classes. The dataset was aggregated from multiple public repositories and expanded through augmentation. The results, authored by Md. Sifat Haque Zidan, were last updated on 2026-04 24.

Use Cases

• Benchmarking tiny vision transformer architectures based on the described validation performance metrics.
• Evaluating custom convolutional neural network designs for efficiency based on the reported parameter counts and accuracy.
• Analyzing model interpretability based on the qualitative Grad-CAM and Grad-CAM++ analyses mentioned in the description.

Strengths

• Results are based on a dataset of approximately 45,000 original images across 51 classes.
• Model performance is quantified with specific validation and test accuracy metrics, such as 99.66% test accuracy for PlantaNet.
• Computational characteristics are detailed, including model sizes (e.g., 9.85 MB) and parameter counts (e.g., 1.28M for PlantaNetLite).

Limitations

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

Provenance

Source

Md. Sifat Haque Zidan

Collection Method

Results from training a classification framework on a dataset aggregated from multiple publicly available repositories.

Time Range

null

Freshness

Last updated 2026-04-24 17:24:13; freshness should be verified.

Geography

null