Performance Comparison of Mainstream Rice Disease Detection Models

5.5 KB of tabular data comparing the performance of the proposed ECL-RTDETR model against a baseline for rice disease detection. The dataset, authored by Yaojun Zhang and last updated on June 3, 2026, likely contains metrics from experiments showing improvements in accuracy, speed, and computational efficiency. The model enhancements include an EfficientViT backbone, CARAFE upsampling, and LDConv in the neck network.

Use Cases

• Benchmarking new object detection models against the reported performance metrics for rice disease detection.
• Evaluating the trade-offs between model accuracy, inference speed, and computational cost in agricultural AI applications.
• Informing model architecture choices based on the reported improvements from using EfficientViT, CARAFE, and LDConv components.

Strengths

• Specific performance improvements are quantified: 0.7% increase in [email protected], 22.2 FPS speed increase, and reductions of 81.8 GFLOPs and 22.12M parameters.
• The technical approach is described in detail, including the use of a lightweight backbone and specialized operators for feature preservation.
• Dataset is openly licensed under CC-BY-4.0.

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, likely containing only summary performance metrics rather than raw image data.

Provenance

Source

Yaojun Zhang, via figshare.

Collection Method

Likely contains experimental results from training and evaluating the ECL-RTDETR model.

Freshness

Last updated 2026-06-03 17:54:34; freshness should be verified.