5.54 GB
89,953 files
Updated 4 months ago
Name
Size
Aircraft Damage.v1-roboflow-instant-2--eval-.coco
Aircraft Damage.v3-roboflow-instant-3--eval-.coco
Aircraft Dent-Crack.v1i.coco
Aircraft Dent-Crack.v2i.coco
Aircraft Dent-Crack.v3i.coco
Aircraft Dent-Crack.v4i.coco
Aircraft Dent-Crack.v5i.coco
Aircraft crack.v2i.coco
Corrosion Detection.v2i.coco
Corrosion Detection.v3i.coco
Corrosion Detection.v4i.coco
Corrosion-detection.v1i.coco
Corrosion-detection.v2i.coco
aircraft_skin_defects.v1-roboflow-2.0-model-stock-2023-04-17-4-46pm.coco
aircraft_skin_defects.v10-isolated-object-4-classes.coco
aircraft_skin_defects.v11-isolated-object-crack.coco
aircraft_skin_defects.v12-isolated-object-dent.coco
aircraft_skin_defects.v13-isolated-object-missing-head.coco
aircraft_skin_defects.v14-isolated-object-scratch.coco
aircraft_skin_defects.v18-3-classes-full-images-aug_gray_flip.coco
aircraft_skin_defects.v19-4-classes-grayscale-full-images-no-aug.coco
aircraft_skin_defects.v20-5-classes-grayscale-no-augmn-full-images.coco
aircraft_skin_defects.v21-5-classes-isolated-grayscale-no-aug.coco
aircraft_skin_defects.v22-single-class-defect-aug-prep.coco
aircraft_skin_defects.v23i.coco
aircraft_skin_defects.v5-yolov5-stock.coco
aircraft_skin_defects.v6-yolov5-single-defect-class.coco
aircraft_skin_defects.v7-roboflow-automl-single-class-defect.coco
aircraft_skin_defects.v8-single-class-defect-with-pre-psg-and-augm.coco
corrosion detection.v1i.coco
crack-hair-dent.v1-roboflow-instant-1--eval-.coco
crack-hair-dent.v2i.coco
crack-hair-dent.v3i.coco
crack-hair-dent.v4i.coco
crack-hair-dent.v5i.coco
crack-hair-dent.v6i.coco
evaluation
images
labels
sample_detections
.gitattributes1.99 kB
xet
README.md4.49 kB
xet
best.onnx7.51 MB
xet
best.pt3.78 MB
xet
best.rknn2.75 MB
xet
best_balanced.onnx7.51 MB
xet
data.yaml212 Bytes
xet
infer_onnx.py4.09 kB
xet
README.md

ZeroDefect — YOLOv5 Nano (v2.0 Class-Balanced)

Aircraft Skin & Automotive Component Defect Detection (Edge-Deployable)

Python PyTorch ONNX License

A YOLOv5 Nano object detection model trained on a class-balanced merged dataset of aircraft skin and automotive component defects. This is version 2.0 of the ZeroDefect model, trained with:

  • AdamW optimizer + Cosine LR decay for stable convergence
  • Class-balanced sampling across 7 defect categories
  • Multi-scale training (random image size between 0.5x–1.5x)
  • Label smoothing (0.05) for better generalization
  • 150 epochs fine-tuning from a high-quality pretrained base

Detected Defect Categories (7 Classes)

ID Class Description
0 crack Structural cracks, fatigue lines, stress fractures
1 dent Impact dents, surface deformations
2 corrosion Rust, chemical wear, oxidation spots
3 scratch Superficial scrapes, paint scratches
4 paint-peel Flaking paint, coating degradation
5 missing-head Missing rivet heads or fasteners
6 defect Generic surface anomalies

Training & Validation Metrics

Run Epochs mAP@0.5 mAP@0.5:0.95 Precision Recall
Balanced v2.0 (this repo, best epoch=101) 150 0.30994 0.22613 0.37777 0.34799
Nano v1.0 (baseline, epoch 149) 150 0.31962 0.23492 0.42944 0.35807

Note: The class-balanced v2.0 run achieves comparable mAP scores to the nano v1.0 baseline while training on a harder, class-balanced dataset. The lower absolute mAP reflects that the dataset had its class distribution enforced (harder to overfit to the dominant "defect" class), resulting in more balanced per-class detection capability.

Training Configuration

  • Model Architecture: YOLOv5 Nano backbone (~1.76M parameters, 4.2 GFLOPs)
  • Dataset: Merged aircraft skin defect + corrosion dataset (~30k images, 7 classes)
  • Optimizer: AdamW | Cosine LR decay
  • Label Smoothing: 0.05
  • Early Stopping: patience=30 (best epoch: 101/150)
  • Hardware: NVIDIA RTX 3050 Laptop GPU (6 GB VRAM)

Repository Contents

File Description Size
best.pt PyTorch model weights (YOLOv5 format) ~3.8 MB
best_balanced.onnx ONNX export (opset 18, simplified) ~7.5 MB
best.rknn RKNN binary for RV1106 NPU (INT8, prior run) ~2.6 MB
data.yaml Dataset configuration <1 KB
infer_onnx.py Python ONNX inference helper ~8 KB
evaluation/ Training curves, confusion matrix, PR/F1 plots —

Quick Start: ONNX Inference (Python)

pip install onnxruntime opencv-python numpy
python infer_onnx.py --model best_balanced.onnx --image test.jpg --conf 0.25

Hardware Deployment (Luckfox Pico Max / RV1106 NPU)

The best.rknn model is compiled for the RV1106 NPU (INT8 quantized):

# Push model to the board
adb push best.rknn /userdata/

# Build C++ inference runner from rknn_model_zoo
./build-linux.sh -t rv1106 -a armv7l -d yolov5

# Run inference on board
./rknn_yolov5_demo /userdata/best.rknn test.jpg

License

Released under the MIT License. See LICENSE for details.

Total size
5.54 GB
Files
89,953
Last updated
Jul 18
Pre-warmed CDN
US EU US EU

Contributors