RESEARCH
AI Generated Characters
Advancements in machine learning have recently enabled the hyper-realistic synthesis of prose, images, audio and video data, in what is referred to as artificial intelligence (AI)-generated media. These techniques offer novel opportunities for creating interactions with digital portrayals of individuals that can inspire and intrigue us. AI-generated portrayals of characters can feature synthesized faces, bodies and voices of anyone, from a fictional character to a historical figure, or even a deceased family member..
Projects
Living Memories: AI-Generated Characters as Digital Mementos
Every human culture has developed practices and rituals associated with remembering those who have died. Our system that allows you to create an “interactive photograph” of a person based on any data source. We ran a study with 90 users interacting with living memories of the three historical figures “Leonardo Da Vinci”, “Murasaki Shikibu”, and “Captain Robert Scott” and found them to increase learning effectiveness and
motivation to learn.Read more.
Keywords:
Generative AI, Large Language Models, Vector Database, Learning, Motivation, Deepfakes.
Publication:
ACM Intelligent User Interfaces.
Role:
Built the system which is one of the first to use vector databases and semantic models together with LLMs. Developed evaluation metrics and, experimental methods, leading write-up.
AI-Generated Virtual Instructors Based on Liked or Admired People Can Improve Motivation and Foster Positive Emotions for Learning
This research presents the results of a study with 134 participants to explore the effects of learning from an AI-generated virtual instructor that resembles a person one likes or admires. We found that they can significantly improve students' motivation towards learning, foster more positive emotions, and boost their appraisal of the AI-generated instructor as serving as an effective instructor. Read more.
Keywords:
Generative AI, Learning, Motivation, Deepfakes.
Publication:
IEEE: Frontiers in Education.
Press:
Hackster.io
Role:
System and experimental design, building generative AI platform, experimental methods, write-up.
AI-Generated Characters for Supporting Personalized Learning and Wellbeing.
Although negative use cases of deepfakes have dominated public conversation, we explore emerging positive use cases of AI-generated characters, specifically in supporting learning and well-being, and make public an easy-to-use pipeline for teachers and healthcare workers to use. Read more.
Keywords:
Generative AI, Learning, Well-being, Deepfakes.
Publication:
Nature: Machine Intelligence 2021.
Press:
“Characters for good, created by artificial intelligence”, MIT News.
“Media Lab Open-sources AI-generated characters library“, Analytics India Magazine.
“AI-Generated Characters are here. They’re just not evenly distributed”, Forbes
“Deepfakes are being used for good: here’s how”, The Conversation
Repository:
GitHub / Google Colab
Role:
Co-initiator, implemented machine learning models in python and made accessible through colab, collaborated in the write-up, and am maintaining GitHub repository.