FashioNova

Thumbnail
Home Page
Prompt Page
Loader page

💡 Inspiration

The COVID-19 pandemic exposed the weaknesses of the fast fashion industry, revealing its unsustainable practices that left over 70 million garment workers without pay. In response, we developed our virtual try-on app to advocate sustainability and equality in fashion. Our platform enables users to explore their personal style while minimizing waste and production. By offering virtual try-ons, we shift the focus from fast fashion to thoughtful consumption. We advocate diversity by encouraging users to experiment with styles from various cultures, fostering an inclusive fashion community.

🔍 What It Does

FashioNova intelligently recommends clothing based on your input in the web app. You can enter any prompt and virtually try on clothes using your camera. Our technology detects your body and overlays the selected garments onto your image, allowing you to see how they would look on you. FashioNova streamlines your shopping experience, saving you time by eliminating the hassle of trying on clothes and searching for outfits. This presents a fantastic opportunity for fashion companies to upload images of their own clothing, allowing customers to virtually "Try before they buy," ultimately enhancing the shopping experience and boosting sales.

⚙️ How we built it

Frontend: We first made our logo with Adobe Express. Then, we created the client side of our web app using React.js and JSX based on a high-fidelity prototype we created using Figma.

Backend: Our backend was programmed in Python with Flask, where it contains the WebRTC video component for the computer vision and VR/AR tech components of the virtual trying on clothes. As well, it has the Cloudflare AI API logic that allows specific clothing recommendations from the existing wardrobe based on the user's prompt.

🚧 Challenges we ran into

We had some issues regarding the WebSocket logic (Socketio server) which caused a massive lag in the real-time video footage with the body node displays and virtual clothes. To avoid this error, we resorted to just displaying the current computer's webcam. We also had trouble navigating the CloudFlare AI API to choose clothing options based on the desired prompt and the machine learning classification dataset we created, but after some hard perseverance, we were able to implement the feature.

✔️ Accomplishments that we're proud of

We are very proud of our ability to create a working video recognition with a body node detection system that allowed us to creatively add the virtual clothes fitted on the user. Creating and training a machine learning model for classification and labelling clothing data was a first for all of us. Also integrating CloudFlare Ai for the first time and integrating it for creative use felt very achievable! Plus, creating a unique idea and having functional backend features and a very artsy front-end design was very rewarding!!!

📚 What we learned

As a team, we shared valuable knowledge and unique experiences, learning from one another throughout the process. We selected video recognition technology for our project, despite it being new to most of us. Although we encountered numerous bugs along the way, we worked together to overcome them, strengthening our collaboration and problem-solving skills.

🔭 What's next for us!

Expanded Try-Ons: We plan to introduce a wider variety of clothing styles and add accessories like shoes, hats, sunglasses, necklaces, and scarves for a more complete virtual experience.
Enhanced Interactivity: We aim to incorporate a location feature that displays the weather forecast and suggests appropriate clothing based on current conditions. As well, as add a 3D rendering/design option to make the clothes look more realistic vibe!
Relevant Datasets: We aim to enhance our model's ability to recognize and understand various clothing styles by utilizing datasets that feature a broader and more diverse range of apparel.
Marketing and future: We hope to partner with clothing companies to add a shopping feature within the app. Clothing stores that want to allow their clothes to be tried on virtually will be added to our shopping catalogue and users can try out clothes without even going to the store. If they like a piece, they will be directed to the company's website link for that product.
Diversity Implementations: Our web app aims to help all kinds of people. We are excited to continue working on this project to add helpful features to people with disabilities/injuries, people living busy lives, and people who want to upgrade their fashion through a cultural and futuristic approach.

We hope to help people have ease with shopping for clothes, by reducing physical strain, and time spent on shopping, and making it more convenient for our future clients. Join us to help improve the future of virtual fashion! Invest in FASHIONOVA !!!