Towards Cross-Lingual Audio Abuse Detection in Low-Resource Settings with Few-Shot Learning

TL;DR

This paper explores cross-lingual audio abuse detection in low-resource languages using pre-trained models and few-shot learning, demonstrating promising generalization and classification capabilities across multiple Indian languages.

Contribution

It introduces a novel approach combining pre-trained audio representations with meta-learning for abusive language detection in low-resource multilingual audio settings.

Findings

Pre-trained models like Wav2Vec and Whisper effectively generalize in low-resource abuse detection.

Few-shot learning with 50-200 samples achieves competitive classification performance.

Feature visualization provides insights into model behaviour and decision boundaries.

Abstract

Online abusive content detection, particularly in low-resource settings and within the audio modality, remains underexplored. We investigate the potential of pre-trained audio representations for detecting abusive language in low-resource languages, in this case, in Indian languages using Few Shot Learning (FSL). Leveraging powerful representations from models such as Wav2Vec and Whisper, we explore cross-lingual abuse detection using the ADIMA dataset with FSL. Our approach integrates these representations within the Model-Agnostic Meta-Learning (MAML) framework to classify abusive language in 10 languages. We experiment with various shot sizes (50-200) evaluating the impact of limited data on performance. Additionally, a feature visualization study was conducted to better understand model behaviour. This study highlights the generalization ability of pre-trained models in low-resource scenarios and offers valuable insights into detecting abusive language in multilingual contexts.