Citizen science website concept using the Watson Tone Analyzer service

IEEE/IBM Watson 2015 Student Showcase

The MiFace citizen science concept website, built using IBM’s Bluemix cloud hosting and the Watson Tone Analyzer service, won a top 5 award in the IEEE/IBM Watson cognitive computing challenge for students. Thanks go to the work of Hansi Mou on the database, and Stephanie Michalowicz for collaborating on data analysis.

MTurk1

MiFace

The work I did as an undergraduate with Dr. Harriet Oster — a former Ph.D. student advised by Paul Ekman — and at VicarVision during the summer of 2013 has evolved into MiFace, a tool for generating recognizable facial expressions using a 3D model. Still in the research stage, my cohort Stephanie Michalowicz and I are hoping to use[…]

Action Unit-driven Avatar

Facial Expression Recognition-driven Avatar

This presentation reviews the work done for my summer 2013 internship at VicarVision, working with their FaceReader automated expression recognition software. A 3D model generated using faceshift was modified and scripted in Maya to conform to the standard Action Units used in coding facial expressions is driven by the data output by FaceReader.

expressionStudy_th

Expressions of Emotion Study

Online Observer Judgement Study Based on Research done by Dr. Harriet Oster This is an interactive expression recognition study I developed  for my Lab in Human Cognition and Advanced Interactive Design classes, using HTML, CSS, JavaScript and jQuery. Data gathering for the original study is complete, but the study is still up if you want to try[…]