Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
library_name: onnx
|
| 4 |
+
pipeline_tag: object-detection
|
| 5 |
+
tags: [object-detection, onnx, yolo11, rt-detr, gemstone, quality-control]
|
| 6 |
+
---
|
| 7 |
+
|
| 8 |
+
# GemScan inclusion detectors (ONNX)
|
| 9 |
+
|
| 10 |
+
YOLO11s and RT-DETR-L fine-tuned to detect inclusions in diamond photos, exported to ONNX
|
| 11 |
+
(opset 17, static 1×3×640×640 input, FP32). `manifest.json` lists each file with its
|
| 12 |
+
architecture, input size and class names.
|
| 13 |
+
|
| 14 |
+
Test split (16 images, 117 inclusions):
|
| 15 |
+
|
| 16 |
+
| Model | Precision | Recall | mAP50 | mAP50-95 |
|
| 17 |
+
|---|---:|---:|---:|---:|
|
| 18 |
+
| YOLO11s | 0.390 | 0.213 | 0.187 | 0.059 |
|
| 19 |
+
| RT-DETR-L | 0.357 | 0.282 | 0.195 | 0.067 |
|
| 20 |
+
|
| 21 |
+
Trained on [Diamond Inclusion](https://universe.roboflow.com/diamond-classification-data/diamond-inclusion) (CC BY 4.0) by Diamond Classification Data.
|
| 22 |
+
Pre/post-processing that reproduces Ultralytics exactly is in `app/inference.py` of the source repository.
|