May 2022

Chinchu Thomas (PH2015017) and Prof. Dinesh Babu Jayagopi, ‘Predicting Presentation Skill of a Speaker Using Automatic Speaker and Audience Measurement’, published in IEEE Transactions on Learning Technologies (Impact factor: 3.7)


Presentation abilities are considered a crucial soft skill for teachers, students, and professionals alike. New ideas and innovative thinking are sparked by effective presentations. In today’s global society, where communication and learning take place both online and offline, presentation skills are vitally important. Automatic assessment of presentation abilities is crucial in this situation for determining presentation quality. Automatic evaluation saves time, money, and human effort, all of which can be saved with manual evaluation. In an effective learning environment, student participation is critical. Teachers and lecturers can use student engagement data to improve and alter their teaching style and instructional tactics based on the audience. Students are more likely to drop out in today’s global context, when learning takes place in a hybrid setting.

In this paper, we suggest a model for predicting a speaker’s presentation skills and student involvement in a classroom situation. To deal with the dataset’s size, we employed domain adaptation. Supervised models were trained using domain adaptation to predict presenting abilities and student engagement. We also looked at the relationship between audience involvement and the speaker’s presentation skills, and discovered that automatic audience engagement prediction and the audience rating of the speaker’s presentation skills are linked. This demonstrates that the automatic audience engagement model can be used as a proxy for evaluating presentation skills.


The International Institute of Information Technology Bangalore is a technical and research university in Bangalore, India. The Institute is a registered not-for-profit society funded jointly by the Government of Karnataka and the IT industry under a public-private partnership model.

For more details: Visit

Featured Stories