IndiSearch

Extension detecting people
IndiTech API search
Try on before (loading)
Try on after

Inspiration

Imagine watching the latest music video from your favorite band or streaming a movie online when a stunning outfit catches your eye. You want to find it online, but tracking it down can be frustrating and time-consuming. That's where IndiSearch comes in.

This project was born from a simple and relatable experience: spotting an outfit you love and having no idea where to find it. Whether it's a scene from a TV show, a post on social media, or a look from a fashion campaign, inspiration can strike anywhere. Unfortunately, access doesn't always follow.

With IndiSearch, we set out to close that gap. By combining computer vision technology with the Inditex API, we created a tool that allows users to search for clothing in any moment, turning style inspiration into real-world discovery.

What it does

Our Chrome extension makes shopping effortless by letting you discover and buy clothing directly from any website. As you browse platforms like Netflix, X or Youtube, the extension automatically detects clothing items within the visuals.

When you hover over or click on a highlighted piece, it instantly searches the Inditex database for similar items you can purchase or virtually try on. Whether it's a jacket from a movie scene or a cap from an influencer's post, the extension helps you find and shop the look in just a few clicks.

Finally you can try this new outfit using the try on feature. Powered by a diffusion model, this tool lets you virtually try on the clothes you are interested in, giving you a more interactive and confident shopping experience.

How we built it

We developed a Chrome extension using JavaScript, supported by a Python backend built with FastAPI. The backend integrates multiple APIs and libraries to enable various core functionalities. To begin with, we used the YOLO model to detect people on the screen and draw bounding boxes around them. For product search, we connected to the Inditex API, allowing users to find clothing items related to the selected visuals. To enhance image data, we integrated SerpApi to perform image scraping of Inditex products. Since the Inditex API requires images to be accessible via public URLs, we also implemented a mechanism to generate temporary image URLs. Lastly, we utilized the Try-On Diffusion endpoint from RapidAPI, allowing users to see how clothes would look on them.

Challenges we ran into

Since we had never worked with Chrome extensions before, developing one took more time than we initially expected. One of the main challenges we faced was with the IndiTech API. It does not return actual product images, only purchase links, so we had to implement an additional step to retrieve the image sources ourselves. We also encountered another issue with the Inditex API, which requires publicly accessible URLs for image inputs. To solve this, we created a method to generate temporary URLs for the screenshots we captured during the browsing experience.

Accomplishments that we're proud of

We developed a lightweight, user-friendly Chrome extension with a non-intrusive design. Despite a tight 36-hour timeline, we successfully overcame every challenge and completed all the tasks and features we had planned from the outset.

What we learned

Throughout this project, we gained valuable experience in a wide range of technologies and techniques. We successfully developed and deployed a fully functional Chrome extension capable of analyzing a user's screen in real-time. We integrated advanced computer vision models like YOLO to detect people in images accurately. We also implemented web scraping tools to retrieve product details, using product names, ids and brands (based on on-screen content and the IndiTech api). These images were then displayed to users in a seamless interface. One of the most exciting parts of our project was using state-of-the-art diffusion models. These models allowed us to take an image of a person and an image of a clothing item, and generate a realistic new image of the person wearing that clothing.

What's next for IndiSearch

Although we are very proud of our application, we recognize that there is always room for improvement. In the long term, we see great potential in enhancing both the accuracy of clothing search and the quality of image generation. With additional time and development, these features could be further refined to offer a smoother and more reliable user experience.

To improve usability even further, there are several features we aim to develop. First, we would like to integrate the extension more seamlessly with Inditex platforms, allowing users to add items directly to their shopping basket without leaving the current page. Another key area for improvement is mobile accessibility. Currently, the Chrome extension is limited to desktop use, but our goal is to make the tool accessible to as many users as possible, including those on mobile devices.