iWAX:可解释的 Wav2vec-AASIST-XGBoost 语音伪造检测框架
Seungeun Lee1,2, Sunmook Choi1,3, Taein Kang4
1Department of Mathematics, Korea University, Seoul, 02841, South Korea.
Scientific reports
|November 18, 2025
概括
本研究介绍了iWAX,这是一个用于检测语音伪造的可解释系统. 它使用深度学习模型 (wav2vec 2.0) 和XGBoost来解释其决策,性能优于现有方法.
科学领域:
- 语音处理 语音处理
- 机器学习 机器学习
- 人工智能的人工智能是人工智能.
背景情况:
- 像 wav2vec 2.0 (w2v2) 这样的深度学习模型越来越多地用于语音伪造检测.
- 然而,它们的复杂性阻碍了解释性,使得很难理解它们的决策过程.
研究的目的:
- 开发一个可解释的语音伪造对策 (iWAX),将w2v2与XGBoost结合起来.
- 通过识别重要的时间和频段来解释检测预测.
主要方法:
- iWAX使用精心调整的w2v2前端和AASIST后端与XGBoost分类器.
- 对于频段分析,应用sinc波器,而时间分析则专注于关键的w2v2特征.
- XGBoost的特征重要性机制是解释性方法的核心.
主要成果:
- 与基线模型 (AASIST,w2v2-AASIST) 相比,iWAX在ASVspoof 2019 LA数据集上表现出优异的性能.
- 该系统为其语音伪造检测预测提供了人类可以理解的解释.
- 使用LightGBM证实了稳定性,表明可解释性方法的广泛适用性.
结论:
- 在语音伪造检测中,iWAX在高性能和可解释性之间取得了强大的平衡.
- 该方法解决了传统和基于深度学习的对策的局限性.
- 这项工作促进了对人工智能驱动的音频安全系统的信任和理解.
相关概念视频
Masking and Demasking Agents
3.4K
EDTA titrations may necessitate masking and demasking agents to temporarily protect a particular metal ion in a mixture from the EDTA reaction. These agents facilitate the sequential analysis of the metal ions by forming stable complexes with some—but not all—metal ions during certain steps.
There are many masking agents, such as cyanide, fluoride, triethanolamine, thiourea, and 2,3-bis(sulfanyl)propan-1-ol (formerly 2,3-dimercapto-1-propanol), with the masking agent chosen based on...
There are many masking agents, such as cyanide, fluoride, triethanolamine, thiourea, and 2,3-bis(sulfanyl)propan-1-ol (formerly 2,3-dimercapto-1-propanol), with the masking agent chosen based on...
3.4K
Impression Management Techniques IV: Altercasting
163
Altercasting is a strategic communication technique in which an individual imposes a specific identity or social role onto another person to influence their behavior and shape the interaction. By presuming a role—such as “responsible leader” or “patient person”—altercasting encourages the target to conform to that identity, often aligning their behavior with the expectations associated with the role. The power of this tactic lies in its subtlety; once a role...
163
Air-entraining Agents
248
Air-entraining agents improve the durability and workability of concrete in climates with frequent freezing and thawing. These agents prevent cracks by introducing small air bubbles into the mix, creating spaces accommodating water expansion when temperatures drop. The air-entraining agents lower the surface tension of water, forming stable, small air bubbles. This method is more effective than having accidental large voids, as the intentional, smaller, and evenly distributed air voids improve...
248
Classification of Signals
1.3K
In signal processing, signals are classified based on various characteristics: continuous-time versus discrete-time, periodic versus aperiodic, analog versus digital, and causal versus noncausal. Each category highlights distinct properties crucial for understanding and manipulating signals.
A continuous-time signal holds a value at every instant in time, representing information seamlessly. In contrast, a discrete-time signal holds values only at specific moments, often denoted as x(n), where...
A continuous-time signal holds a value at every instant in time, representing information seamlessly. In contrast, a discrete-time signal holds values only at specific moments, often denoted as x(n), where...
1.3K
Perceiving Loudness, Pitch, and Location
905
The human brain perceives pitch through two primary mechanisms reflected in place theory and frequency theory. Each mechanism describes how sound waves are interpreted as specific pitches by the brain, offering insights into the intricate processes of auditory perception.
Place theory, or place coding, suggests that different pitches are heard because various sound waves activate specific locations along the cochlea's basilar membrane. The brain determines the pitch of a sound by...
Place theory, or place coding, suggests that different pitches are heard because various sound waves activate specific locations along the cochlea's basilar membrane. The brain determines the pitch of a sound by...
905
Perception of Sound Waves
5.4K
The human ear is not equally sensitive to all frequencies in the audible range. It may perceive sound waves with the same pressure but different frequencies as having different loudness. Moreover, the perception of sound waves depends on the health of an individual's ears, which decays with age. The health of one's ears may also be affected by regular exposure to loud noises.
The pitch of a sound depends on the frequency and the pressure amplitude of the source. Two sounds of the same...
The pitch of a sound depends on the frequency and the pressure amplitude of the source. Two sounds of the same...
5.4K


