# Lightweight Toxicity Detection in Spoken Language: A Transformer-based Approach for Edge Devices

arXiv: https://arxiv.org/abs/2304.11408 · HTML: https://mlsanity.com/paper/2304.11408/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Knowledge distillation for LLMs, one of 12 directions in Knowledge distillation.
- Builds on 19 of its 58 references that are in the map.
- First posted on arXiv 22 Apr 2023.

## Placed in

- Knowledge distillation / [Knowledge distillation for LLMs](https://mlsanity.com/distill/knowledge-distillation-for-llms/)

## Builds on (most cited first, 19 of 19)

- [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://mlsanity.com/paper/2006.11477/) (2020)
- [Common Voice: A Massively-Multilingual Speech Corpus](https://mlsanity.com/paper/1912.06670/) (2019)
- [MELD: A Multimodal Multi-Party Dataset for Emotion Recognition in Conversations](https://mlsanity.com/paper/1810.02508/) (2018)
- [A Survey of Model Compression and Acceleration for Deep Neural Networks](https://mlsanity.com/paper/1710.09282/) (2017)
- [Mobile-Former: Bridging MobileNet and Transformer](https://mlsanity.com/paper/2108.05895/) (2021)
- [A Comparative Study on Transformer vs RNN in Speech Applications](https://mlsanity.com/paper/1909.06317/) (2019)
- [Dawn of the transformer era in speech emotion recognition: closing the valence gap](https://mlsanity.com/paper/2203.07378/) (2022)
- [Compressing Large-Scale Transformer-Based Models: A Case Study on BERT](https://mlsanity.com/paper/2002.11985/) (2020)
- [The fifth 'CHiME' Speech Separation and Recognition Challenge: Dataset, task and baselines](https://mlsanity.com/paper/1803.10609/) (2018)
- [Transformer-based Acoustic Modeling for Hybrid Speech Recognition](https://mlsanity.com/paper/1910.09799/) (2019)
- [Multi-Task Learning in Natural Language Processing: An Overview](https://mlsanity.com/paper/2109.09138/) (2021)
- [Handling Bias in Toxic Speech Detection: A Survey](https://mlsanity.com/paper/2202.00126/) (2022)
- [Exploring Wav2vec 2.0 fine-tuning for improved speech emotion recognition](https://mlsanity.com/paper/2110.06309/) (2021)
- [MulT: An End-to-End Multitask Learning Transformer](https://mlsanity.com/paper/2205.08303/) (2022)
- [Fine-tuning wav2vec2 for speaker recognition](https://mlsanity.com/paper/2109.15053/) (2021)
- [Transformers in Speech Processing: A Survey](https://mlsanity.com/paper/2303.11607/) (2023)
- [Iterative Distillation for Better Uncertainty Estimates in Multitask Emotion Recognition](https://mlsanity.com/paper/2108.04228/) (2021)
- [DeToxy: A Large-Scale Multimodal Dataset for Toxicity Classification in Spoken Utterances](https://mlsanity.com/paper/2110.07592/) (2021)
- [Emotion Detection in Text: a Review](https://mlsanity.com/paper/1806.00674/) (2018)

## Built on by (most cited first, 5 of 5)

- [Adversarial Fine-Tuning of Language Models: An Iterative Optimisation Approach for the Generation and Detection of Problematic Content](https://mlsanity.com/paper/2308.13768/) (2023)
- [Enhancing Multilingual Voice Toxicity Detection with Speech-Text Alignment](https://mlsanity.com/paper/2406.10325/) (2024)
- [Towards Probing Speech-Specific Risks in Large Multimodal Models: A Taxonomy, Benchmark, and Insights](https://mlsanity.com/paper/2406.17430/) (2024)
- [ViToSA: Audio-Based Toxic Spans Detection on Vietnamese Speech Utterances](https://mlsanity.com/paper/2506.00636/) (2025)
- [Beyond Content: A Comprehensive Speech Toxicity Dataset and Detection Framework Incorporating Paralinguistic Cues](https://mlsanity.com/paper/2605.15984/) (2026)

## Abstract (arXiv)

Toxicity is a prevalent social behavior that involves the use of hate speech, offensive language, bullying, and abusive speech. While text-based approaches for toxicity detection are common, there is limited research on processing speech signals in the physical world. Detecting toxicity in the physical world is challenging due to the difficulty of integrating AI-capable computers into the environment. We propose a lightweight transformer model based on wav2vec2.0 and optimize it using techniques such as quantization and knowledge distillation. Our model uses multitask learning and achieves an average macro F1-score of 90.3% and a weighted accuracy of 88%, outperforming state-of-the-art methods on DeToxy-B and a public dataset. Our results show that quantization reduces the model size by almost 4 times and RAM usage by 3.3%, with only a 1% F1 score decrease. Knowledge distillation reduces the model size by 3.7 times, RAM usage by 1.9, and inference time by 2 times, but decreases accuracy by 8%. Combining both techniques reduces the model size by 14.6 times and RAM usage by around 4.3 times, with a two-fold inference time improvement. Our compact model is the first end-to-end speech-based toxicity detection model based on a lightweight transformer model suitable for deployment in physical spaces. The results show its feasibility for toxicity detection on edge devices in real-world environments.