Knowledge Distillation for Real-Time Classification of Early Media in Voice Communications

TL;DR

This paper presents a novel, resource-efficient approach using knowledge distillation and gradient-boosted trees for real-time classification of early media in voice calls, achieving faster performance with comparable accuracy.

Contribution

It introduces a low-resource, gradient-boosted tree-based method with knowledge distillation for early media classification, outperforming CNN-based models in speed while maintaining accuracy.

Findings

Significant runtime improvement over CNN models.

Comparable classification accuracy with the new approach.

Effective performance in a regional data center case study.

Abstract

This paper investigates the industrial setting of real-time classification of early media exchanged during the initialization phase of voice calls. We explore the application of state-of-the-art audio tagging models and highlight some limitations when applied to the classification of early media. While most existing approaches leverage convolutional neural networks, we propose a novel approach for low-resource requirements based on gradient-boosted trees. Our approach not only demonstrates a substantial improvement in runtime performance, but also exhibits a comparable accuracy. We show that leveraging knowledge distillation and class aggregation techniques to train a simpler and smaller model accelerates the classification of early media in voice calls. We provide a detailed analysis of the results on a proprietary and publicly available dataset, regarding accuracy and runtime performance. We additionally report a case study of the achieved performance improvements at a regional data center in India.