Domain.com Submission: photofy.online (Photo + Spotify = Photofy)
photofy gives users a more uniquely personal way to interact with your photos, by listening to the messages they send. Some see mountains and seek an adventure, others seek calmness, our web app allows you to truly explore the places you visit in a unique way.
Two of our team members flew from India for this hackathon, and they did what most toursits do, they took photos of the landmarks the environment around them. It got us thinking, we gaze at the photos we take when we get back home and we try to re live the photo. But now, with our Web app you can get music based on the location, environment and sentiment of the users in the photo, all curated into a Spotify playlist.
What it does:
photofy allows anyone to upload an image and effecitvely 'hear' the area or relive the joy. For example, when you look at a picture of a mountain some see calmness, others see an adventure, and our app curates the songs based on the user by allowing them to connect their Facebook profile. Once the image is uploaded our smart backend curates a playlist based on the image and your interests.
How we built it:
Our web app uses the express.js framework (and node.js). We primarily utilize the Google Cloud Platform, in which we use the Vision API to detect Sentiment, landmarks and other characteristics from photos, Google Cloud Storage buckets for storing the photos on our servers as we process them, the Google Maps GeoCoding API to decode the metadata from photos and find where they are in the world and lastly Google's Compute engine to host our website as well as backup everything just for fun. We also used AWS to host the photos to pass to Alexa for future development. All of the above linked with Spotify's API to create the playlists based on keywords, and some parsed lyrics.
Challenges we ran into
We started hacking late, our main challenge was how to work together given our diverse technical backgrounds. We ended up attempting a project none of us have a lot of experience with. Our backgrounds included Voice, Machine Learning and Algorithm design. And so we challenged ourselves to learn and try a new type of hack and work together.
Accomplishments that we're proud of
We're very proud that though we didn't solidify an idea until there were 12 hours left, we still persevered and put together a demoable product. We're really happy with how it turned out, and we're excited to improve on it and mayeb have it a regular feature of Spotify in the future.
What's next for photofy
photofy will hopefully be a base product that comes with Spotify one day, allowing for an even more personal feel to the music you listen to.