Recent transplants love to take pride in how much they've assimilated to the city's culture. Well, here's your chance to quantify how cool and cynical you've become by comparing your opinions with those of all other New Yorkers—at least, those active on Twitter in 2018.
The computer will volley hot topic keywords from the past year, and two players will have one shot to prove how aligned they are with most common New York opinions. The closer your response is to the topic, the more points you score— but you have to be quick, because nothing is less cool than trying too hard.
My project allows the user to generate a new song title and lyrics based on the Alan Lomax collection of American folk songs and 19th century song lyrics through an interactive interface. The code works based on LSTM machine learning model. I trained an LSTM model based on the song titles and lyrics of the Alan Lomax's collection available on the Library of the Congress website. The project generates song titles and I am working on training the model with lyrics. I would like to generate images based on the text in the next steps for this project.
More on the collection: Alan Lomax and other members of the Lomax family are associated with an enormous number of “classic” folk songs and traditional tunes. This list contains some of most widely-known and frequently performed “iconic” pieces that the Lomaxes documented through their field recordings, or increased public awareness of through commercial recordings, publications, radio programs, and concerts.
You are now standing at virtual forest. Take a moment to look around and listen. All sounds here are the LATENT REPRESENTATION of real forests. You are listening to the meaningful features of forest sound encoded and decoded through latent space with deep neural networks.
Generative Music, Machine Learning for the Web, The Neural Aesthetic