Beardom

Inspiration

Facial hair has been an integral part of human culture and customs since prehistoric times. What started as a somewhat effective method to keep warm during the winters now is a social phenomenon, whether through religion, fashion or simply something to be proud of. Beards, mustaches and other forms of facial hair have become a symbol of masculinity, maturity, and sophistication all around the world.

Our team notices, however, a shortage of support and recommendation apps in facial hair without extensively long tests and visits to the doctor, and as such, decided to make a facial feature recognition web-app that would suggest products and styles for the user. Furthermore, we also recognised the subjectivity of recommendations by ‘professionals’ to be highly subjective and sometimes inaccurate! As such, we aim to deliver the most accurate suggestion by implementing feedback from real people around the world.

What it does

Beardom is an app that will provide you expert beard recommendations, styling and care advice specifically tailored towards your preferences all from the comfort of your own home. The software measures key dimensions of the face, such as face width, jawline width, distance from cheek bone to cheek bone, and face length etc etc… In order to determine the precise shape of your face and hence provide recommendations on specific beards that will suit your preferred style. In addition to styling suggestions, our app will also provide step by step instructions and timelines to grow your desired beard. During this process the app will recommend relevant styling and care products to ensure your hair is healthy and looking fresh every single day!

How we built it

The project mainly consists of two main parts : Facial feature landmark detection and the recommendation system:

Facial Feature landmark detection was carried out using the mediapipe library by google.

We used the content based recommendation system to create the recommendations for the products. We used the TFIDF vectorizer and the created the sparse matrix. The comparison scores were brought down to 0 to 1 range using the sigmoid function.

Challenges we ran into

Time: Lack of Coding Experience: overcame it by coding the program not in the most efficient way

Accomplishments that we're proud of

-Teamwork -Coding knowledge we gained -Developing a fully functioning code -Successfully completing the checkpoints on time

What we learned

-Teamwork -Patience -Coding experience -For most of us this was the first every large coding project that we did. We are proud that we successfully completed it

What's next for Beardom

This software and concept and be expanded upon other recommendation systems, including but not limited to: Hairstyles - The same facial recognition features existing in your app, such as the face size dimension recordings , could be applied hair style suggestions Earrings, Makeup and Accessories - while this app has been initially catered towards young/middle aged men, the software and idea could be extended to cater for women as well by utilising a moisture and inflammation detection feature to including suggestions for skin care routine and beauty products to keep the skin healthy. Rating feature - Enables other users from around the world to anonymously rate beard styles in criterias such as: attractiveness, manageability, groomability, cleanliness etc… This subsequently enables us to improve our database and hence allowing the software to improve recommendations tailored to your interests. Smart detection using genetic information - Enables user to input their genetic information when creating their profile to allow more precise tracking of hair growth and provide timely product recommendations.