The UCL Condita team selected for the 2021 Alexa Prize TaskBot Challenge
4 August 2021
A team of UCL students is among the 10 teams selected to compete in this year’s prestigious Alexa Prize TaskBot Challenge. This year-long competition is the first conversational AI challenge to incorporate multimodal (voice and vision) customer experiences.
A team of UCL students is among the 10 teams selected, from over 125 universities worldwide, to compete in this year’s prestigious Alexa Prize TaskBot Challenge. This is a year-long competition and the first conversational AI challenge to incorporate multimodal (voice and vision) customer experiences. This challenge consists of building a TaskBot to help customers complete complex tasks in two domains, cooking and DIY. The UCL team is named Condita, which stands for “COoking aNd DIw Task-based Agent”, has received $250,000 of funding from Amazon for their participation. Condita in Italian means “seasoned” when written as “condita” or “with fingers” when written as “con dita”.
The team is led by doctoral student Yue Feng as the Team Lead and Dr Aldo Lipani as the Faculty Advisor. Yue is a doctoral student in the Dept of Computer Science (CS) advised by Prof. Emine Yilmaz and Dr Lipani. Dr Lipani is a Lecturer (Assist. Prof.) in Machine Learning in the Dept of Civil, Environmental and Geomatic Engineering (CEGE) and a member of the SpaceTimeLab and Web Intelligence group. The rest of the team consists of UCL students from the CEGE and CS depts.
The other selected universities for this year’s Alexa Prize Challenge include Carnegie Mellon University, National Taiwan University, NOVA School of Science and Technology, Ohio State University, Texas A&M University, University of California Santa Barbara, University of Glasgow, University of Massachusetts Amherst, and the University of Pennsylvania. The winning team will receive a $500,000 prize.
The Condita team’s focus will be on the IR and NLP aspects of the challenge on cooking and do-it-yourself (DIY) domains. The team’s vision is to provide an enjoyable user experience with professional step-by-step instructions on Cooking and DIY tasks by leveraging a rich task-knowledge base, multi-modal interactions, user personalisation, and human-like fluency.
The Condita team is excited to take part in the Alexa Prize TaskBot Challenge. “We have recently designed methods to evaluate conversational systems and developed models for task-oriented dialogue systems, and this challenge gives us a unique opportunity to test our models within the Alexa platform by interacting with real users,” concludes Dr Lipani.