SkinScan is a model and app that I developed to predict certain skin diseases, these being chickenpox, measles, and shingles. The project was completed during my last semester as CSU Chico, for my capstone class. Below is an outline of the folders contained in this repo. All code is public use.
Contains the images that I scraped from multiple different search engines to train, validate, and test the model. If you are interested in using this dataset for your own training, I have made it open to use. You can also download the dataset from my website apache server with the following link. http://capstone.colehopkins.me/dataset/
contains my model developed on Google Colab using Tensorflow, Keras, Pandas, Sklearn, and other tools for graphing and data collection. Python code and A IPython Notebook Document. You can also reach my code directly from Google Collab with the following link. https://colab.research.google.com/drive/1lI5O5yeaUz2JPTYllkuMqU2J6bFGbVxa?usp=sharing
contains my React Native Code, which is still incomplete, that uses the resulting model and makes predictions given an image. I used expo during my development process, and was able to get some great results, although I did not reach my requirement to display the predictions.
Contains my draft proposal, draft requirements and test plans, and my design and development update. Also contains my presentation slides given during my capstone presentation.
The poster image was used during my poster session, and is currently hanging up in the Computer Science Department. It contains useful information about what my project is, what it does, and the progress I made.
Thank you so much for checking out my project, if you have any questions, feel free to reach out to me at coleh52012@gmail.com.