# Explaining deep learning models for spoofing and deepfake detection with SHapley Additive exPlanations

arXiv: https://arxiv.org/abs/2110.03309 · HTML: https://mlsanity.com/paper/2110.03309/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Oct 2021.

## Built on by (most cited first, 13 of 13)

- [How Do Neural Spoofing Countermeasures Detect Partially Spoofed Audio?](https://mlsanity.com/paper/2406.02483/) (2024)
- [Audio Deepfake Detection Based on a Combination of F0 Information and Real Plus Imaginary Spectrogram Features](https://mlsanity.com/paper/2208.01214/) (2022)
- [FairSSD: Understanding Bias in Synthetic Speech Detectors](https://mlsanity.com/paper/2404.10989/) (2024)
- [SLIM: Style-Linguistics Mismatch Model for Generalized Audio Deepfake Detection](https://mlsanity.com/paper/2407.18517/) (2024)
- [DSVAE: Interpretable Disentangled Representation for Synthetic Speech Detection](https://mlsanity.com/paper/2304.03323/) (2023)
- [What Does an Audio Deepfake Detector Focus on? A Study in the Time Domain](https://mlsanity.com/paper/2501.13887/) (2025)
- [Protect-Your-IP: Scalable Source-Tracing and Attribution against Personalized Generation](https://mlsanity.com/paper/2405.16596/) (2024)
- [FSBI: Deepfakes Detection with Frequency Enhanced Self-Blended Images](https://mlsanity.com/paper/2406.08625/) (2024)
- [A Data-Driven Diffusion-based Approach for Audio Deepfake Explanations](https://mlsanity.com/paper/2506.03425/) (2025)
- [An Explainable Proxy Model for Multiabel Audio Segmentation](https://mlsanity.com/paper/2401.08268/) (2024)
- [ATLAS: AI-Native Receiver Test-and-Measurement by Leveraging AI-Guided Search](https://mlsanity.com/paper/2508.12204/) (2025)
- [Sparse deepfake detection promotes better disentanglement](https://mlsanity.com/paper/2510.05696/) (2025)
- [What Do Deepfake Speech Detectors Actually Hear?](https://mlsanity.com/paper/2606.10912/) (2026)

## Abstract (arXiv)

Substantial progress in spoofing and deepfake detection has been made in recent years. Nonetheless, the community has yet to make notable inroads in providing an explanation for how a classifier produces its output. The dominance of black box spoofing detection solutions is at further odds with the drive toward trustworthy, explainable artificial intelligence. This paper describes our use of SHapley Additive exPlanations (SHAP) to gain new insights in spoofing detection. We demonstrate use of the tool in revealing unexpected classifier behaviour, the artefacts that contribute most to classifier outputs and differences in the behaviour of competing spoofing detection models. The tool is both efficient and flexible, being readily applicable to a host of different architecture models in addition to related, different applications. All results reported in the paper are reproducible using open-source software.