App for skin cancer detection - Udacity Deep Learning NanoDegree 2020
App for skin cancer detection - Udacity Deep Learning NanoDegree 2020
source code: https://github.com/umair13adil/tensorflow_lite_flutter
project: https://drive.google.com/open?id=1ZkKHbpX6LVj_XERUBjImDkIBzN55e5fS (My Google Drive)
Notes: Small datasets tend to overestimate class prediction results in image classification tasks. As the image shown (app.png), the biggest accuracy is the first model option when there are low camera resolution, low light and bad framing. When app starts, the biggest accuracy will be also shown first.
Result1.png: Taken from Internet, it’s a melanoma example. The model also points to the biggest accuracy (seborrheic keratosis).
Result2.png: Taken from Internet, it’s a nevus example. The model also points to the biggest accuracy (seborrheic keratosis).
APK: https://drive.google.com/open?id=1ce9jh7p1L2AZUQtXoSf0A5UcZxgAZX6W