Description

This project uses deep learning to classify product images as either defective (def_front) or non-defective (ok_front). It leverages a pretrained ResNet-18 model. The tool includes scripts for training the model and predicting defects from new images.

---

Pre-requisites

Python 3.8 or higher

---

Create a Python Virtual Environment

python -m venv venv venv\Scripts\activate

---

Install the required packages

pip install torch torchvision matplotlib pillow

---

Train the model

python scripts/train.py

---

Test the model

python scripts/predict.py