Declutterly Classifier – v1.0

🧠 Declutterly Classifier – Pretrained Model

This is the pretrained model file (gallery_model.pth) for the Declutterly App, a neural network-based tool that classifies images from a typical phone gallery into one of the following categories:

• 🐾 Animals
• 🍔 Food
• 😂 Memes
• 👤 People
• 🌇 Places
• 📱 Screenshots

---

✅ What’s included

• MobileNetV2 base (pretrained on ImageNet)
• Fine-tuned on a custom dataset of ~10,000 images scraped from Reddit
• Lightweight classifier head with 6 output classes
• Input resolution: 224 × 224 × 3

---

🚀 How to use

1. Clone the repository.
2. Run the app script.
3. If gallery_model.pth is not found locally, the script will automatically download this model.

No need to retrain the model unless you want to improve or adapt it further.

---

📌 Notes

• This model was trained using transfer learning with frozen base layers.
• Optimized for CPU performance and small academic projects.