Muvi: making photos alive

Logo

Inspiration

We wanted to build an App based on AI that could be used in real life and provided a real value to the users. Inspired by our last trip to Basilea, we came up with the idea of making a project that could help you remember the good times that you spent through your photos

What it does

Our model is a text-to-video project, although we do not use a generative video model. Instead, we use the user's photos together with ai music generation to create a video. The user can upload photos, and the model will generate a video with the photos and music. The music is generated using an ai model, and the video is created using the photo and the music. The user can download the video and share it with friends.

How we built it

We used a lot of different tools, many of them based on AI. We also used a Vectorial DB provided by InterSystems (IrisVectorial).

Challenges we ran into

Connecting so many different models and dealing with the uncertainty that comes from the fact of using AI in your code proved to be a great challenge, but a loot of trial and error and good coding practices ended up in a working project.

Accomplishments that we're proud of

At first we thought that the initial idea was to ambitious to be developed in a weekend. After 36 intense coding hours, we are proud to deliver a project that makes even more than what we first imagined.

What we learned

Working in a team of four people, developing a big project part by part and dividing the workload to maximize productivity and develop an incredibly ambitious project

What's next for Muvi: making photos alive

Adding metadata to the vectorial DB to let users select certain dates or places in order to provide a more personalized experience when creating a video. Also, improve the user interface in order to make task like looking at your gallery or deleting photos easier for the user.