A Multimodal Machine Learning Framework for Teacher Vocal Delivery Evaluation

TL;DR

This paper introduces a multimodal machine learning framework that objectively evaluates teacher vocal delivery by analyzing fluency and passion, addressing subjectivity and inefficiency in manual assessments.

Contribution

The study presents a novel pairwise comparison method with a multimodal orthogonal fusion algorithm for large-scale, objective vocal delivery evaluation.

Findings

Effective evaluation of vocal delivery achieved

Datasets collected from real-world education scenarios

Code made publicly available

Abstract

The quality of vocal delivery is one of the key indicators for evaluating teacher enthusiasm, which has been widely accepted to be connected to the overall course qualities. However, existing evaluation for vocal delivery is mainly conducted with manual ratings, which faces two core challenges: subjectivity and time-consuming. In this paper, we present a novel machine learning approach that utilizes pairwise comparisons and a multimodal orthogonal fusing algorithm to generate large-scale objective evaluation results of the teacher vocal delivery in terms of fluency and passion. We collect two datasets from real-world education scenarios and the experiment results demonstrate the effectiveness of our algorithm. To encourage reproducible results, we make our code public available at \url{https://github.com/tal-ai/ML4VocalDelivery.git}.