EYEBEAM EXCHANGE | Learning AI: Creating Emotional Data Sets and Deep Learning in the Browser with ML5
Learning AI: Creating Emotional Datasets and Deep Learning in the Browser with ML5
Saturday, April 21, 2018
Doors open at 10:30am
11:30 am – 5pm
This event is now full. Watch the live stream here:
We welcome donations to continue supporting technology by artists.
Session 1: Led by Hannah Davis
Session 2: Led by Cristóbal Valenzuela & Yining Shi
We kick off our first month of public programs in Eyebeam’s new home with a focus on Artificial Intelligence. This month’s workshop is inspired by the work of current resident Stephanie Dinkins. While in residence at Eyebeam, Stephanie’s focus is to create a unique culturally attuned AI entity in collaboration with coders, engineers and in close consultation with local communities of color that reflects and is empowered to work toward the goals of its community.
The day will consist of two artist-led workshops that explore the tools used for deep learning and the datasets that can be created by them. Workshops will begin with an introduction on the potentials of culturally attuned AI by Stephanie Dinkins.
Our goal with the Exchange is to introduce curious learners to new subjects through tools, context and examples, providing enough insight so that students feel more comfortable exploring these topics after the workshop. You will walk away with more questions then you came in with and we hope you are as excited about that as we are.
Admission to Eyebeam Exchange covers a full day of learning and includes participation in both workshops!
Please bring your own laptop to the class.
Space is limited, BUY TICKETS HERE!
Session 1: Emotional / Subjective Data and Dataset Creation, Led by Hannah Davis
11:30 am – 1:30 pm
This session will look at emotional/subjective data and dataset creation, we will be coding in Python. We will explore what ‘subjective data’ means, how to capture abstractions like emotions and other subjective experiences in datasets and examine currently available emotional datasets, exploring them critically.
Participants will learn to identify subjectivity in “objective” datasets by looking at how the datasets were created, who tagged them, who funded the dataset and what the dataset leaves out. We will explore “artisanal data”, creating explicitly subjective datasets for artistic purposes and talk about how a culture’s values can be a type of bias. This workshop will explore how to avoid bias retention over time including experiments with labels and vocabulary – how we can create these, and use them with regularity. Finally, participants will create and explore their own subjective dataset with a focus on how to be cognizant during the collection, analysis, and usage stages.
Basic programming knowledge, particularly in Python, would be helpful, but this will be an introductory session aimed at beginners.
Hannah Davis is a programmer and generative musician from NYC. Her work falls along the lines of music generation, data sonification, natural language processing, and sentiment analysis. Her algorithm TransProse, which translates novels and other large works of text into music, has been written up in TIME, Popular Science, Wired, and others. A human-computer collaboration, where she analyzed the sentiment of articles talking about technology over time, was performed by an orchestra at The Louvre this past fall. Hannah is currently working on creating unique datasets for art and machine learning, and is also working on a project to generativly score films. She is a 2017 AI Grant recipient. see also musicfromtext.com
Session 2: Deep Learn Web, Led by Cristóbal Valenzuela & Yining Shi w/ Special Guest
2:30 pm – 4:30 pm
Cristóbal Valenzuela Cristóbal Valenzuela is a technologist and software developer interested in building digital tools and interactive experiences. His work has been sponsored by Google and the Processing Foundation. He is currently a M.P.S Candidates at the Interactive Telecommunications Program (ITP).
Yining Shi is a creative technologist and a software engineer. Her research interests lie in developing novel ways of learning and teaching computational topics through various media like machine learning, creative coding, physical computing, and data visualization. Yining is also a contributor to various open source projects from the Processing Foundation and New York University’s Interactive Telecommunications Program.