https://github.com/sans11pentium/garbage-classification-chsw
Classify an image into the various classes of garbage.
https://github.com/sans11pentium/garbage-classification-chsw
streamlit
Last synced: about 2 months ago
JSON representation
Classify an image into the various classes of garbage.
- Host: GitHub
- URL: https://github.com/sans11pentium/garbage-classification-chsw
- Owner: Sans11Pentium
- Created: 2025-04-12T03:47:30.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2025-04-25T04:11:35.000Z (about 1 year ago)
- Last Synced: 2025-06-23T20:50:26.833Z (about 1 year ago)
- Topics: streamlit
- Language: Jupyter Notebook
- Homepage: https://garbage-classifier-chsw.streamlit.app/
- Size: 19.2 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 🗑️ Garbage Classifier Streamlit App
This is a simple web application built with Streamlit that allows users to classify an uploaded image of waste into one of six predefined categories. It utilizes a trained model in TFLite format to make real-time predictions directly in the browser.
## 🚀 Features
- Upload an image of waste material
- Automatically predicts the class of garbage
- Easy-to-use and interactive interface
- Fast and efficient inference using TensorFlow Lite
### 🛠️ MobileNet to TFLite Conversion for Edge Impulse (TinyML Deployment)
To deploy the trained MobileNet model on an edge device using **Edge Impulse**, it was converted to the TensorFlow Lite (TFLite) format. This enabled real-time inference on resource-constrained hardware as a TinyML application.
Link to live inferencing on EdgeImpulse : https://smartphone.edgeimpulse.com/classifier.html
#### Steps Followed:
1. **Model Training**
This lightweight `MobileNetV2` model using TensorFlow/Keras trained on the garbage classification dataset is designed to be efficient and suitable for edge deployment.
2. **TFLite Conversion**
After training, model was converted to a `.tflite` format using the TFLite converter.
## 🧠 Supported Classes
The app classifies the uploaded image into one of the following categories:
1. **Glass**
2. **Paper**
3. **Cardboard**
4. **Plastic**
5. **Metal**
6. **Trash**
## 📦 Access the app
Link is provided in the desription of this repository.