The problem FootWellness: Your Foot Doctor solves

• FootWellness is a web application that uses computer vision and machine learning algorithms to analyze an image of a person's foot and determine the type of arch in their feet - flat, normal, or high. This information can be used to recommend appropriate footwear and exercises to improve foot health and prevent potential foot-related injuries.

• The application allows users to upload an image of their foot and get a quick analysis of their arch type. It is especially helpful for those who suffer from foot pain, have a history of foot injuries, or just want to make sure they are wearing the right shoes for their feet.

• With the help of FootWellness, people can easily identify their foot arch type and get personalized recommendations to keep their feet healthy and pain-free. The application saves time, effort, and money by avoiding unnecessary visits to a foot specialist and purchasing incorrect footwear.

Challenges we ran into

• Data collection: Collecting a sufficient amount of high-quality data can be a challenge, especially if the dataset is specific to a certain demographic or population group.

• Class imbalance: The distribution of classes in the dataset may not be balanced, which can result in poor performance of the classifier for minority classes.

• Feature selection: Identifying the most informative features to use in the classification model can be challenging. It may require domain expertise and feature engineering.

• Model selection: Selecting an appropriate machine learning model that can accurately classify foot arch types can be challenging, as different models have different strengths and weaknesses.

• Overfitting: Overfitting can occur if the model is too complex and fits the training data too closely, resulting in poor generalization performance on new data.

• Interpretability: Understanding how the model makes its predictions and which features are important can be important for medical applications, but some machine learning models, such as deep neural networks, can be difficult to interpret.

• Proper Contouring