gphua1/rklb-defect-model

RKLB Component Defect Detection Model

Model Description

This model is designed for automated quality control in manufacturing, specifically for detecting defects in components.

• Task: Binary Image Classification (Normal vs Defective)
• Architecture: efficient_vit
• Input Size: 224x224 RGB images
• Classes: Normal, Defective
• Accuracy: 97.5%

Usage

With the RKLB Defect Detection Space

The easiest way to use this model is through the RKLB Materials Space.

Programmatic Usage

from huggingface_hub import hf_hub_download
import torch

# Download model
model_path = hf_hub_download(
    repo_id="gphua1/rklb-defect-model",
    filename="best_model.pth"
)

# Load model
checkpoint = torch.load(model_path, map_location='cpu')
# ... initialize your model architecture and load weights

Training Details

• Framework: PyTorch
• Model Type: Vision Transformer (ViT) variant
• Training Data: Manufacturing component images
• Task: Binary classification for quality control

Intended Use

This model is intended for:

• Automated quality inspection in manufacturing
• Component defect detection
• Production line quality control
• Training data augmentation for quality systems

Limitations

• Designed for specific component types
• Best performance on similar lighting conditions as training data
• Binary classification only (normal/defective)

Citation

If you use this model, please cite:

@misc{rklb-defect-model,
  author = {Gary Phua},
  title = {RKLB Component Defect Detection Model},
  year = {2024},
  publisher = {HuggingFace},
  url = {https://huggingface.co/gphua1/rklb-defect-model}
}