CNN Performance Comparison for Toxic Herbal Image Classification

A 5.5 KB Excel file contains performance metrics for a comparison of convolutional neural network architectures on a toxic herbal image classification task. The dataset, created by Guohua Zhu and last updated in March 2026, reports that EfficientNetV2 achieved the highest performance across all evaluated metrics. The specific number of model comparisons and the underlying image dataset size are not provided.

Use Cases

• Benchmarking CNN architectures based on reported performance metrics for image classification.
• Selecting a model for plant toxicity identification based on comparative evaluation results.
• Analyzing the relationship between model architecture and classification performance on herbal imagery.

Strengths

• File size is precisely 5.5 KB, indicating a compact, focused results table.
• Includes a clear performance benchmark, stating EfficientNetV2 achieved the highest metrics.
• Released under a permissive CC-BY-4.0 license for open reuse.

Limitations

• Row count and the number of compared architectures are unknown.
• Column-level documentation is absent; field semantics must be inferred after download.
• The underlying image dataset used for evaluation is not described or included.

Provenance

Source

Guohua Zhu via figshare

Collection Method

Likely contains experimental results from model training and evaluation.

Freshness

Last updated 2026-03-19 20:02:55; freshness should be verified.