This repository contains our notebooks and dataset that we used when exploring our dataset, building our model, and testing our middleware. These notebooks was run on Google Colab environment.
We used trashnet by Gary Thung. You can get the original one from here.
We have our two versions of that dataset, which is v1 and v2. Both of this dataset originally from dataset-resized.zip. v1 is the version where we manually classify and merged the "trash" class to the other classes. While v2 is the version with "trash" class entirely removed from the dataset.
 Run in Google Colab
|
 View source on GitHub
|
 Download notebook
|
For all of process that we do, we are using Google Colab (you can choose on the link above). But you can use local instead by several steps:
-
Download notebook
-
Clone the project [Optional]
$ git clone https://github.com/GWaste/waste-classifer
-
Go to the project directory
$ cd waste-classifer -
Install the required library with virtualenv
- Linux/ MacOS
$ virtualenv env $ source env/bin/activate (venv) $ pip install numpy matplotlib tensorflow - Windows
python -m venv env env\scripts\activate pip install numpy matplotlib tensorflow
- Linux/ MacOS
-
Run with jupyter notebook
$ jupyter notebook
Contributions are always welcome!
Feel free to clone, use, and contribute via pull request.
Got an issue? Please use issues panel
We are exciting to see your contributions!
If you have any feedback, please reach out to us at [email protected] or contact one of our member.
Here are some related paper that we used:
Fine-Tuning Models Comparisons on Garbage Classification for Recyclability
Comparative Analysis of Multiple Deep CNN Models for Waste Classification