Mpowerr HECMS YOLOV8S

Shape Leaf Left - Mpowerr HECMS
Shape Leaf Right - Mpowerr HECMS
About Image - Mpowerr HECMS

Mpowerr HECMS YOLOV8S — Balanced Speed and Accuracy for Real-World Applications

This model strikes an excellent balance between compact size, fast inference, and high accuracy. It is well-suited for real-world object detection scenarios, delivering consistent performance across both edge devices and cloud-based systems.

Performance Metrics — Accuracy, Insights, Outputs

This section highlights the core metrics used to evaluate the model’s performance, including precision, recall, and mAP scores. It also features evaluation charts and inference outputs that demonstrate how the model performs across various conditions.

96.00%

Precision

01

92.45%

Recall

02

97.05%

mAP@0.5

03

76.41%

mAP@0.5:0.95

04
F1 Confidence Curve
Precision Confidence Curve
Recall Confidence Curve
Precision Recall Curve
Confusion Matrix
Normalized Confusion Matrix
Training & Validation Metrics
Label Distribution Heatmap
Label Correlogram
Model Output 1
Model Output 2
Model Output 3
Model Output 4
Model Output 5
Model Output 6
Model Output 7
Model Output 8
Model Output 9
Model Output 10
Model Output 11
Model Output 12

Changelog

V1

Mpowerr HECMS – YOLOV8S V1 Release

  • First stable release of the YOLOV8S model fine-tuned for elephant detection in Sri Lankan environments.
  • Achieves 96.00% precision, 92.45% recall, and 97.05% mAP@0.5.
  • Comprehensive evaluation includes PR curves, F1-score curves, confidence plots, and detection heatmaps.
  • Open-source model weights and performance reports are available for community use and testing.
  • Balanced for both accuracy and speed, making it suitable for deployment on edge devices and cloud APIs.