A deep learning application to help you decide whether your household products belong in the trash, recycle, compost or e-waste.
Confused if that egg-carton belongs in the trash or recycle? Not sure what to do with your dead batteries? Classify all you household waste by simply taking or uploading a picture of it on my website/app.
Under the hood, predictions are being made by a classification model that builds upon the classic MobileNetV2 model.
See the model in production on HuggingFace Spaces: https://huggingface.co/spaces/chaninder/SmartWaste
Here is the model itself to be used for open-source purposes: https://huggingface.co/chaninder/waste-sorting-model
Here's how the app looks:
HOME:
Take a picture or upload one!
After you've given the app a picture, it predicts the category of waste!
- TensorFlow
- TensorFlow Lite
- Transfer Learning
- Git
- Flutter
- Gradio
- Google Collab
- Jupyter Notebook
- Visual Studio Code
- Getting a Flutter developer onboard to create a more asthetic UI/interface for the app
- Image Segmentation: Two new features to incorporate: (1) User can take a picture of MULTIPLE items (rather than just one). The model will output the proper label for each image. (2) Real-time labeling - User takes a video and the app labels each item in real time.