NCSI: Quantify Beauty with AI

Status	Completed
Mentor Name	Liwen Shih
Mentor's XSEDE Affiliation	XSEDE Campus Champion & inaugural Fellow
Mentor Has Been in XSEDE Community	4-5 years
Project Title	Quantify Beauty with AI
Summary	Technology progress into higher definition (HD) demands high-quality content strategy. This project aims in automating the creating and discerning aesthetic levels among visual contents of images, art, graphics, etc... (and can be extended later beyond visual content). Data will be extracted from measuring/filtering/transforming visual contents for AI machine learning analysis to uncover the mystery of what constitute absolute/biological/emotional beauty and harmony, thus enabling the synthesis, selection and creation of beautiful contents for our HD world.
Job Description	1. Learn and program image processing. 2. Learn to program to extract data from measuring/filtering/transforming visual contents. 3. Explore data visualization and machine learning to form hypothesis for artificial beauty rules. 4. Work as XSEDE student campus champions and teaching/equipment assistants for monthly XSEDE training.
Computational Resources	1. XSEDE supercomputing Clusters with Data Mining. 2. Data visualization software. 3. AI & Machine Learning software.
Contribution to Community
Position Type	Apprentice
Training Plan	Training Approach – Building a Machine Learning Model that takes an image and classifies whether it is conforms to a (pre-programmed) standard of beauty, pertaining to the nature of the image. Training Format - weekly meetings + self-guided tutorials/papers/websites + hands-on programming Student learning goals - Intro to Machine Learning and Artificial Intelligence, Learn Python, TensorFlow, Keras, Spark Training resources: 1. Course – Udemy Python 3 Bootcamp, Google ML Crash Course, Udacity TensorFlow 2.0 Course, Microsoft Artificial Intelligence Track 2. Website – Google, Udemy, Udacity, YouTube, Microsoft 3. Reading – Textbook: Hands-On Machine Learning with Scikit-Learn and TensorFlow, Supplemental: The Golden Ratio: The Divine Beauty of Mathematics 3/18/2019 Golden Ratio Book (for possible applications) 5 hrs Selection of type of ML Model and Domain 3/25/2019 Google ML Course 10 hrs Primitive ML Model 4/1/2019 XSEDE Big Data & ML Training 10 hrs Beginner Big Data Badge 4/8/2019 Hands on ML with TensorFlow 10 hrs More Robust, trained Model 4/15/2019 Final Poster & Report 10 hrs Successful Poster Session at Student Conference
Student Prerequisites/Conditions/Qualifications	1. Interest and background in visual contents. 2. Thrive on challenge in complex problem solving. 3. Team synergy. 4. Effective communication.
Duration	Semester
Start Date	01/22/2019
End Date	05/06/2019