README.md

🔌 Circuit AI – Circuit Classification Web App

A deep learning web application built with PyTorch + FastAPI that can train, evaluate, and predict electronic circuits (Amplifier, RC Highpass, RC Lowpass, Resonator, Other) directly from a user-friendly dashboard , calculates the values of the passive elements and creates a netlist.

---

🚀 Features

• Dashboard-style UI with plain HTML/CSS/JS
• User Authentication (Login & Register)
• Upload Images for circuit prediction
• Live Training with real-time loss & accuracy per epoch
• Organized dataset structure for easy training/testing
• Passive elements values calculation
• Netlist creation

---

📂 Project Structure

CircuitImageAi/
│── app.py                # FastAPI backend
│── static/
│   ├── images.jpg        # Background image
│   ├── login.html        # Login & Register UI
│   ├── style.css         # Styling
│   └── script.js         # Frontend logic
│── templates/
│   ├── login.html        # Login UI
│   ├── dashboard.html    # Dashboard UI
│   └── register.html     # Register UI
│── dataset/
│   ├── train/
│      ├── amplifier/
│      ├── rc_lp/
│      ├── rc_hp/
│      ├── resonator/
│      └── other/
│── models/               # Here are saved some previous trainings
│── images/               # Some images for testing 
│── requirements.txt      # Dependencies
└── README.md             # Documentation

---

⚙️ Installation

1. Clone the repo

git clone https://github.com/your-username/CircuitImageAi.git
cd CircuitImageAi

2. Install dependencies

pip install -r requirements.txt

---

▶️ Running the App

Start the FastAPI server:

uvicorn app:app --reload

Open your browser:
👉 http://127.0.0.1:8000/static/login.html

---

📊 Dataset

Organize your dataset as follows:

dataset/
├── train/
   ├── amplifier/   # training images
   ├── rc_lp/
   ├── rc_hp/
   ├── resonator/
   └── other/

⚠️ Place at least 20–30 images per class in train/ and 2-5 images per class in test/ for decent results. Insisde the helpers folder there are to subfolders with some datasets , one big and one small.

---

Configure the Database

The credentilas are stored in a json file that is called users.json .

🧑‍💻 Usage

1. Register/Login

   • Register a new account or login with existing credentials.
   • <img width="1920" height="1041" alt="Image" src="https://github.com/user-attachments/assets/e0d334a6-8d8a-4b2d-a67e-7bde5030dab5" />
   • Or login with existing credentials.
   • <img width="1920" height="1043" alt="Image" src="https://github.com/user-attachments/assets/b235ebe1-f45a-40f9-8283-c0f6900661e0" />

2. Training

   • Start training or continue a previous one from the dashboard.
   • <img width="1920" height="913" alt="Image" src="https://github.com/user-attachments/assets/df13d68a-2253-47af-950b-78438c638c82" />
   • Monitor loss & accuracy per epoch in real time.
   • Or you can load a previous training.

3. Prediction

   • Upload an image of a circuit.

   • Get the predicted class + confidence percentage instantly.

   • In case of filter :

   • Enter the cutoff frequency.

   • Click Compute & plot

   • Fill the Library and cell name

   • Click Download Netlist

   • <img width="1920" height="1080" alt="Image" src="https://github.com/user-attachments/assets/dc77afee-6a49-4f2a-aa05-4172d82a2ba6" />

   • In case of resonator

   • Choose vowel

   • Click Load & plot

   • Fill the Library & Cell name

   • Click DownLoad Netlist

   • <img width="1920" height="1080" alt="Image" src="https://github.com/user-attachments/assets/a79bd09d-fb58-41dc-8579-96dc753604c9" />

4. Settings

   • Here the User can change his email, phone number or add a new password
   • <img width="1920" height="1080" alt="Image" src="https://github.com/user-attachments/assets/35afeb26-29dc-49fe-8580-0af216eb5550" />

---

📌 Requirements

Add these to requirements.txt:

fastapi
uvicorn
torch
torchvision
pillow
python-multipart
passlib[bcrypt]
cryptography

---

✅ To-Do

• [ ] Add more circuits

---