https://github.com/kevinlu1248/recycler
Recyclables manager fine-tuned on XCeption using Kaggle's GPU instances.
https://github.com/kevinlu1248/recycler
deep-learning google-automl recycler xception
Last synced: about 1 year ago
JSON representation
Recyclables manager fine-tuned on XCeption using Kaggle's GPU instances.
- Host: GitHub
- URL: https://github.com/kevinlu1248/recycler
- Owner: kevinlu1248
- Created: 2020-03-30T00:12:35.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2022-11-22T02:42:42.000Z (over 3 years ago)
- Last Synced: 2025-02-04T21:41:56.278Z (over 1 year ago)
- Topics: deep-learning, google-automl, recycler, xception
- Language: Python
- Homepage: https://recycler-cv.herokuapp.com/
- Size: 1.05 MB
- Stars: 2
- Watchers: 2
- Forks: 0
- Open Issues: 8
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Recycler
For sorting recyclables via image recognition. Try it out at https://recycler-cv.herokuapp.com/.
## Process:
I started by directly using Google Vision ML API but it gave information that was too vague to use to classify the recyclables.
It gave me information such as "bottle", or "products", so I tried fine-tuning on Xception, which was a lot more accurate
(86% precision). Used Gary Thung's dataset at https://github.com/garythung/trashnet. I am planning also to
add more from Huawei's garbage sorting competition https://competition.huaweicloud.com/information/1000007620/circumstances?track=107,
which has around 20000 images, although the majority of them do notapply to this project.