📅 Workshop Schedule (50 min per session)

The workshop is divided into five structured sessions. This allows participants to gradually understand the concepts and have enough time to test each step without feeling overwhelmed.

---

Session 1: Introduction & Data Collection

📌 Objective: Explain the project and capture images for training the model.

• Brief introduction to CNNs, IoT, and computer vision.
• Explanation of the training and inference process.
• Setting up the Raspberry Pi camera.
• Capturing images, organized by category (at least 20 per class).
• Uploading images to a PC or Google Drive for training.

💡 By the end of this session, participants will have a dataset ready for preprocessing.

---

Session 2: Preprocessing & Training the Model

📌 Objective: Train a simple CNN model to classify objects.

• Brief explanation of how CNNs work.
• Using Jupyter Notebook / Google Colab to prepare data:
  • Resizing and normalizing images.
  • Creating training and validation sets.
• Building and training a CNN model using TensorFlow/Keras.
• Evaluating the model and fine-tuning hyperparameters.

💡 By the end of this session, participants will have a trained model ready for deployment.

---

Session 3: Model Conversion & Raspberry Pi Setup

📌 Objective: Convert the trained model to TensorFlow Lite and prepare it for execution on Raspberry Pi.

• Converting the model to .tflite for optimized edge device performance.
• Transferring the model to the Raspberry Pi.
• Installing necessary dependencies on Raspberry Pi (TensorFlow Lite, OpenCV).
• Testing the model to ensure it loads correctly on the Raspberry Pi.

💡 By the end of this session, Raspberry Pi will be ready for real-time inference.

---

Session 4: Real-time Object Classification

📌 Objective: Use Raspberry Pi to classify objects in real-time using the camera.

• Explanation of how inference works with TensorFlow Lite.
• Capturing a real-time image using the Raspberry Pi camera.
• Preprocessing the image (resizing and normalizing).
• Running inference with the trained .tflite model.
• Displaying the classification result on the screen using OpenCV.

💡 By the end of this session, participants will see live object classification in action.

---

Session 5: 3D Designing & Printing a Custom Mount

📌 Objective: Design and 3D print a custom stand for the Raspberry Pi and camera.

• Quick explanation of why a stable mount improves classification accuracy.
• Designing in FreeCAD / Fusion 360 / Tinkercad:
  • Stable base for the Raspberry Pi.
  • Adjustable support for the camera.
  • Ventilation for heat dissipation (optional).
• Preparing and 3D printing the mount.
• Assembling the hardware into the printed mount.

💡 By the end of this session, participants will have a custom 3D-printed mount for their project.