The problem SkinFit solves

It can identify upto 10 most common skin disorder including acne, eczema, keratosis, warts etc.
The solution can rival the accuracy of doctor’s diagnosis
It tends to provide immediate help to the patients in case of emergency
It brings healthcare to those patients who do not visit doctors for the reasons:
1.They cannot afford specialist treatment
2.Embarrassment or downplaying severity of the symptoms

It has following features:

1. Suggests immediate cure along with credible medical remedies
2. Recommends best certified doctor with a facility to contact them
3. Finds the nearby chemist using current location
4. Handles daily reminders according to follow-up schedule
5. Provides a skin plan for after-treatment precautions

It intends to add the following features in future:

1. The accuracy of the model can be improved by adding more training images and tuning the model architecture.
2. Doctors recommendations can be made on the basis of the patient’s priorities and prerequisites
3. Tracks daily recovery of the disease through regular image upload
4. Skin disease prediction through webcam for direct data entry without the need to upload the image
5. The application can become a platform to connect doctors and patients online

Challenges we ran into

1. We lack GPUs in our system due to which machine learning program execution becomes too slow. To solve this problem, we use online google colabs.
2. In google colabs, we were not able to access the data from our local directory and it was a large file too,therefore, it took a long time to upload it online. To solve this problem, we convert it to zip file so that it can be uploaded soon.
3. We were only two members in the team and none of us do had the time to create a web application, therefore, we use online website creating tool to create a dummy web application.
4. Our created web scraper was unable to work on dermnet website properly, therefore to fix this, we used image downloader extension and prepare the entire data from scratch using the downloaded raw images.