AP Comparison Across Multiple IoU Thresholds for Blood Cell Detection

A dataset from 2026 contains performance metrics for the Elastic YOLO object detection model on peripheral blood smear images. The model achieved a [email protected] of 94.7% and [email protected]:0.95 of 87.8% on clinician-annotated data, outperforming baseline models. The data, shared by Neha Margret Issac on figshare, supports research in automated hematology workflows.

Use Cases

• Benchmark object detection models for blood cell recognition based on reported mAP scores.
• Evaluate model performance across different Intersection over Union thresholds for medical imaging tasks.
• Support the development of automated systems for peripheral blood smear analysis mentioned in the description.
• Aid in research on adaptive convolutional networks for morphology-aware detection of cells.

Strengths

• Model performance metrics are explicitly reported, including [email protected] of 94.7% and [email protected]:0.95 of 87.8%.
• Data is shared under a permissive CC-BY-4.0 license.
• The underlying dataset was annotated by clinicians, suggesting expert validation.

Limitations

• The dataset is very small at 5.5 KB, indicating limited scope, likely containing only summary performance metrics.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for detailed analysis.

Provenance

Source

figshare

Collection Method

Likely contains aggregated performance metrics from model evaluation experiments.

Freshness

Last updated 2026-05-12 17:44:41; freshness should be verified.