---
license: mit
datasets:
- shahzaibvohra/realwaste
metrics:
- accuracy
base_model:
- google/efficientnet-b0
pipeline_tag: image-classification
---


# Trash Optimizer Model Repository

## Project Repository

Trash Optimizer repository is hosted on github: https://github.com/cpoisson/trash-optimizer

## Dataset

This project contains a collection of models designed for optimizing waste management through image classification. The models are trained on the following dataset:

- **RealWaste**: [joebeachcapital/realwaste](https://www.kaggle.com/datasets/joebeachcapital/realwaste) - UCI Machine Learning Repository ([DOI](https://archive.ics.uci.edu/dataset/908/realwaste))
- **Recyclable and Household Waste Classification**: [alistairking/recyclable-and-household-waste-classification](https://www.kaggle.com/datasets/alistairking/recyclable-and-household-waste-classification)
- Custom dataset for specialized categories (batteries, light bulbs, electronics, etc.)

## Dataset Preprocessing
The datasets have been preprocessed and assembled using a custom tool available in the `dataset` directory of the repository.

## Training scripts

The training scripts for the models are available in the `scripts` directory of the repository. Each script is designed to train a specific model architecture on the provided datasets.