Papers for
voice security teams
Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.
Voice recognition models evaluated against human perception of cloned voices
VIPBench: A Human-Aligned Benchmark for Voice Identity Perception in the Age of Voice Cloning
Abstract: Speaker-recognition benchmarks test how accurately a model recovers the speaker label of a recording, treating voice identity as a property of the person who produced it. However, these benchmarks do not evaluate how humans perceive voice identity. Whether a listener hears two voices as the same person has consequences for intellectual property rights, cybersecurity, and everyday life. We introduce the Voice Identity Perception benchmark (VIPBench), which contains 124,876 same-or-different judgments from 1,290 listeners on 9,800 recording pairs from 100 public figures. The pairs include real recordings, voice clones generated by a text-to-speech system, and morphs between two speakers' voices. We also define four tasks that evaluate the alignment between speaker-representation models and human identity perception. The first two tasks measure how well a model's similarity score predicts the proportion of listeners who respond "same" and the majority response. The other two measure whether a model and listeners find the same speaker pairs similar, and whether a model's relationship to listener judgments holds equally for real and synthetic recordings. We evaluate eleven publicly available speech models. Listener judgments diverge from speaker labels most often for voice clones: in 41% of responses, listeners judged a clone and a real recording of the same speaker to be different people. No model reaches the noise ceiling; the best model explains 61% of the explainable variance. Models trained with speaker labels align with listeners better than self-supervised models do, yet a model with a 12% error rate on speaker labels aligns with listeners as well as models with error rates below 2%. VIPBench enables the evaluation of speaker models against human perception on real, cloned, and morphed speech, and our results motivate training speaker models directly on listener judgments.
Replayfactor improves detection of replayed deepfake speech attacks
ReplayFactor: Replay-Aware Dual-Attribute Learning for Practical Speech Deepfake Detection
Abstract: Advances in synthesis techniques have made deepfake speech increasingly realistic, amplifying voice security threats, including unauthorized voice cloning and impersonation. Although existing deepfake detectors perform well on digitally generated speech, their performance often degrades on physically replayed deepfakes, which constitute a low-cost and practical threat to real-world systems. Existing research focuses primarily on dataset construction and evaluation, while available dataset diversity and replay-aware detector design remain limited. We propose ReplayFactor, a dual-attribute framework that enhances replay evidence by fusing self-supervised learning (SSL) representations with an acoustic prior, while leveraging multi-task learning and supervised contrastive learning to promote robust, generalizable representations. We also construct ReplayFake-CN, a large-scale, multi-device Chinese dataset comprising approximately 185k replayed synthetic utterances to complement public benchmarks. Our framework improves macro-F1 from 60.99% to 76.44% and reduces EER from 11.86% to 3.05% on the EchoFake open set over the baseline, demonstrating strong potential for practical replayed deepfake detection.
Speech tokens enable realistic voice imitation with chosen text
From Speech Tokens to Stolen Voices: Attacker-Chosen-Text Speaker Impersonation via Representation Inversion
Abstract: Speech interaction systems have gradually evolved from traditional cascaded systems that use text as intermediate information to end-to-end speech language models that directly process speech representations. In some applications, the speech frontend runs on the user's device and sends speech representations to a remote server for inference, while the original audio remains local. However, these representations may still contain speaker identity information. If an attacker obtains them, it may use this information to imitate the user's voice, threatening their voice privacy. To investigate this risk, we propose \stia, a Speech Token Inversion Attack that generates speech with attacker-specified text in a victim's voice using a single short leaked representation, without access to the original audio or its transcript. Specifically, \stia converts discrete, continuous, or mixed speech representations into conditioning sequences and trains a conditional flow-matching speech generation model using pairs of different utterances from the same speaker in public speech datasets. The model combines speaker characteristics contained in the leaked representation with the specified text to generate speech in the victim's voice. Experiments show that, with a Moshi representation leak corresponding to 3 seconds of speech, \stia achieves an average CAM++ speaker cosine similarity of 0.78 and a target-text word error rate of 3.80\%. These results show that speech frontend representations reveal reusable speaker identity information and highlight the need to consider privacy risks when designing speech frontend representation methods.
WavLM layers reveal how speech models detect fake audio attacks
Interpreting SSL Representations for Spoof Detection: a WavLM Study
Abstract: Self-supervised speech models such as WavLM achieve strong spoofing detection performance, yet which layers encode the most spoof-relevant information and what acoustic properties they capture remain opaque. We present a layer-wise interpretability study of WavLM using the ASVspoof 5 database, combining three complementary analyses: (i) per-layer linear probing with attack-level evaluation across speech synthesis, voice conversion, and adversarial attacks; (ii) a canonical correlation analysis (CCA) linking layer representations to spoofing-relevant acoustic features; (iii) an exhaustive grid search over layer subsets showing that use of five layers outperforms full thirteen-layer pooling. CCA analysis reveals that, while most acoustic features peak in alignment with mid-layers and then decay, voice quality measures jitter, shimmer, and the harmonics-to-noise ratio show the opposite trend, with correlation increasing toward later layers. Embedding visualisations reveal that deeper layers capture a continuous acoustic similarity among attacks rather than discrete categories, suggesting these layers encode fundamental synthesis artifacts rather than attack-specific signatures.
Quantum circuits improve fairness in deepfake speech detection confidence
QBFF: Deep Quantum Bottleneck for Automatic Speaker Verification Anti-Spoofing
Abstract: A central goal of quantum machine learning is to identify applications in which small quantum circuits can improve task performance while remaining executable on real hardware. We investigate deepfake (synthetic) speech detection, a largely unexplored application in this area, where reliable confidence scores are as important as discrimination performance because deployment decisions are made at fixed thresholds and overconfident errors can be costly. We introduce QBFF, which routes a compressed speech representation through a small 8-qubit variational circuit (VQC) used as a drop-in bottleneck, with a matched classical bottleneck as the control. The quantum circuit yields better-calibrated confidence in most matched comparisons, including out-of-distribution evaluation, though the downstream deployment-cost benefit is model-dependent. We then run the same circuit on a real IBM quantum processor. To our knowledge, this is the first anti-spoofing pipeline to run on a QPU; hardware execution demonstrates feasibility, with lower subset EER and $C_{llr}$ but higher mean actDCF than the matched classical reference. Together, these results show that hardware-executable quantum bottlenecks can improve score reliability, with gains that vary by evaluation setting.
TTS speech watermarks improve traceability under signal changes
GAMW: Generation-Aware Mel-Domain Watermarking for TTS Traceability
Abstract: As text-to-speech (TTS) systems continue to advance, provenance tracking and security verification of synthesized speech have become increasingly important. Audio watermarking provides active provenance by embedding imperceptible, machine-recoverable information into speech. However, existing post-hoc methods typically modify the waveform after synthesis, leaving the watermark vulnerable to compound perturbations. To address this limitation, we propose GAMW, a watermark injection framework integrated into the TTS generation process. Before waveform synthesis, GAMW embeds a content-adaptive multi-bit watermark residual into the generated representation, and its decoding objective combines focal modulation with dynamic per-bit weighting. We train and evaluate GAMW separately on two TTS backbones. With a 16-bit payload, GAMW improves decoding accuracy over the strongest baseline by 9.2 and 23.5 percentage points under individual and compound perturbations, respectively, while maintaining competitive speech quality. Reliable clean-speech decoding is also retained at higher payloads. Code is available at https://github.com/ZConJugaTe/GAMW.
Speaker verification systems tested across diverse real-world scenarios
OpenSVBench: Diagnosing Speaker Verification Generalization Across Deployment Scenarios
Abstract: Speaker verification (SV) systems now achieve very low error rates on standard public protocols, but those results do not fully describe deployment generalization. VoxCeleb-style evaluations remain useful anchors for English in-the-wild media speech. Scenario-specific evaluations and challenges are also valuable. However, neither setting directly shows whether the same system remains reliable across diverse deployment scenarios. We introduce OpenSVBench, a public benchmark for cross-scenario SV evaluation. Its fixed public core contains 26 trial sets organized into 11 scenario groups, with 9.15M verification trials, frozen target/non-target keys, and metadata for scenario-wise reporting. Around this core, the online leaderboard acts as a diagnostic evaluation interface rather than only a ranked list. It accepts score packages from new systems, evaluates them with frozen keys and common scoring, and reports trial-set breakdowns, scenario-group metrics, macro EER/minDCF, model-wise profiles, and scenario-wise comparisons. We evaluate 18 public checkpoints spanning x-vector, ECAPA-TDNN, CAMPPlus, ResNet and ERes variants, ReDimNet2, SAM-ResNet, and pretrained Transformer speaker recipes. The results show clear cross-scenario gaps: systems with similar macro EER can have different scenario profiles, strong performance in one scenario does not guarantee strong performance in another, and same-scenario trial sets can still expose source and protocol gaps.
Adaptive purification strengthens audio deepfake detection against attacks
Detection-Guided Adaptive Purification with Diffusion Models for Robust Audio Deepfake Detection
Abstract: Audio deepfake detectors remain vulnerable to adversarial perturbations that suppress the acoustic cues used for detection, allowing manipulated utterances to evade the detector. Although existing defenses can improve robustness, they require retraining the detector or introduce additional distortion. Diffusion-based purification instead leaves the pretrained detector unchanged, but existing methods use the same purification strength for all inputs, creating a trade-off between removing adversarial perturbations and preserving the subtle spoofing cues needed for detection. In this paper, we propose Detection-Guided Adaptive Purification (DGAP), a diffusion-based defense that adjusts purification strength per input. Building on the observation that a light purification perturbs the detector score of an adversarial input far more than that of a benign one, the framework uses the resulting score shift as a reference-free indicator of adversarial manipulation. Inputs with small shifts are passed unchanged, whereas flagged inputs undergo stronger purification before final detection. We evaluate the framework against three adversarial attack settings across three deepfake detectors, and compare it with nine existing defenses. Our results show that DGAP achieves the best defense performance across all detectors while leaving benign inputs nearly unaffected, and remains effective under the defense-aware adaptive attack.
Acoustic and content attacks impact multilingual voice anonymization effectiveness
Exploiting Acoustic and Content-Oriented Speaker Verification Attacks Against Multilingual Voice Anonymization
Abstract: Attacker ASV systems for voice anonymization have been studied primarily in English, leaving their behavior in multilingual settings largely unexplored. Conventional ASV has shown that both acoustic and contextual information are important for multilingual speaker verification. Inspired by this, we investigate whether the same holds for attacker ASV on anonymized speech. We evaluate both acoustic- and content-oriented attackers on multilingual anonymized speech and construct a multilingual voice-converted dataset to improve cross-lingual generalization. Our results show that attacker effectiveness depends on the linguistic utility of the anonymized speech. Overall, acoustic-oriented attackers achieve better performance. However, when linguistic information is well preserved, the performance gap between content- and acoustic-oriented attackers narrows compared with conditions involving stronger speech distortion. The multilingual voice-converted dataset further improves performance and partially reduces the cross-lingual gap. These findings highlight the need for more comprehensive attacker modeling and evaluation protocols that consider both privacy and utility, rather than relying on a attacker strategy\footnote{Full code and pretrained models and MultiVC Dataset link are available at: https://github.com/monkeyDarefeen/DAST
Partial speech spoofing detection improves false alarm control but challenge remains
Temporal Anchors and Editing Sensitivity in Partial Speech Spoofing: A Controlled Study
Abstract: Partial-spoof detectors must reject synthetic content while accepting benign edits. We diagnose temporal anchors and boundary-consistency training using frozen WavLM features, nonlinguistic unit controls, and source-locked evaluation. Genuine-genuine and genuine-fake splices contrast editing false alarms with synthetic-content misses; they do not isolate a unique causal artifact. With layer-6 features, phone units have higher PartialSpoof evaluation frame EER than soft frames (11.18% versus 9.59%). In 500 retrospective PS-eval cases, augmentation reduces genuine-splice false alarms; synthetic-core misses increase at 5% PS-development FPR but not demonstrably at 1%. A 200-case Llama A follow-up retains the false-alarm reduction but does not establish a miss-rate increase. Constructed-case ranking can improve while fixed-threshold misses rise. Consistency gives no uniform gain, and encoder-layer rankings change across corpora. We report frame and event localization with three seeds, including word units, and treat PS evaluation as retrospective. Audit code, selected training scripts, and result summaries are available at https://github.com/mysxs/partial-spoof-diagnostics.
Speech deepfake detector improves spotting fake voices on new sources
GLAD: Global-Local Adaptive Detector for Robust Speech Deepfake Detection
Abstract: Recent advances in AI-based speech synthesis have enabled highly realistic speech, increasing the importance of speech deepfake detection (SDD) in preventing misuse. While mainstream Self-Supervised Learning (SSL)-based detectors achieve strong performance, they suffer from poor generalization to unseen domains and often overlook fine-grained signal artifacts due to a bias towards global semantic consistency. In this paper, we conduct the first detailed empirical and visual analysis to validate these limitations explicitly. Our investigation reveals two critical architectural vulnerabilities: (1) a systemic failure to capture localized spoofing traces, and (2) a severe lack of adaptability to domain-driven shifts in SSL layer importance, rendering static aggregation strategies prone to overfitting. To address these vulnerabilities, we propose the Global-Local Adaptive Detector (GLAD). Specifically, to capture localized forgeries, GLAD employs a Hierarchical Global-Local (HGL) backbone that explicitly bridges the granularity gap by fusing global linguistic and acoustic features with fine-grained local signal details. To counter layer importance shifts in out-of-distribution (OOD) scenarios, we introduce a Hierarchical Adaptive Gating (HAG) mechanism that dynamically recalibrates layer-wise focus in a sample-specific manner. Finally, to address shortcut learning induced by environmental biases, we introduce SaniBoost, a composite data augmentation strategy for robust signal standardization and noise sanitization. Extensive experiments demonstrate that GLAD significantly outperforms state-of-the-art methods, particularly on unseen domain cases.The code will be released upon publication.
Speech deepfake detection improves by fixing training conflicts
DGS-MLDG: Domain Gradient Surgery Guided Meta-Learning for Domain Generalization in Speech Deepfake Detection
Abstract: Speech deepfake detection faces significant challenges due to domain shifts. Domain generalization (DG), particularly meta-learning for domain generalization (MLDG), offers a promising solution by simulating and mitigating domain shifts. However, MLDG is often hindered by conflicting gradients between its meta-train and meta-test objectives, leading to suboptimal performance. To address this problem, we propose domain gradient surgery (DGS), a meta-learning method that resolves conflicts through an asymmetric projection strategy. DGS removes the destructive component from the meta-test gradient, ensuring a conflict-free optimization trajectory versus the meta-train gradient. Furthermore, we introduce layer-wise DGS (LW-DGS), an efficient variant of DGS that dynamically identifies and intervenes only conflict-prone layers. Extensive experiments on challenging benchmarks demonstrate that DGS-MLDG and LW-DGS-MLDG achieve an average relative EER reduction of 5.29% and 4.04%, respectively.
Speech deepfake detection improves using combined acoustic features
What Survives the Codec Shift: Pooled No-Vocals Residuals for Speech Deepfake Detection
Abstract: The transition from vocoder-based to neural-codec speech synthesis makes generalization more difficult for speech deepfake detectors, particularly those relying on speech-oriented representations. It remains unclear which acoustic representations retain discriminative information when the generation mechanism changes. We therefore compare 12 acoustic representations using a shared low-capacity linear classifier to identify effective evidence under codec shift. The analysis shows that hierarchical XLS-R leads on the pooled test set, while pooled no-vocals residual statistics perform best on the unseen-codec condition, revealing complementary behavior across generation conditions. Building on this finding, we propose MN-P, a dual-view detector that integrates an utterance-level pooled no-vocals representation with token-level XLS-R features through adaptive gating. The proposed MN-P reduces EER by 54.2% overall and by 60.9% on the codec-unseen condition relative to the best-performing retrained state-of-the-art system, with consistent gains across different detector backends. These results indicate that pooled no-vocals residual statistics provide effective complementary evidence for cross-generation speech deepfake detection.
Graph-based wavelet tool cuts parameters for speech deepfake detection
WST-Graph: Topology-Preserving Wavelet Scattering Front-End for Speech Deepfake Detection
Abstract: The acoustic front-end determines which forensic cues a speech deepfake detector can exploit. The wavelet scattering transform (WST) provides stable multiscale coefficients with explicit coordinates, yet direct flattening obscures the parent relation between paths. We introduce WST-Graph, reconstructing these paths as a sparse modulation-carrier grid for an AASIST graph backend. Modulation-level normalization and length-aware adaptive local attention pooling produce fixed relative-time representations while retaining the acoustic axes before learned adaptation. This yields a waveform-to-graph interface with a fixed, parameter-free WST. Our configurations remain competitive with AASIST while using approximately 60% fewer trainable parameters and show clear gains on selected out-of-domain benchmarks. These results underscore the value of preserving parent-child relations within the carrier-modulation topology when constructing a compact, physically grounded interface for graph-based speech deepfake detection. Code will be released at https://github.com/saki-ciallo/wst-graph.
Audio deepfake detection improves across languages using language orthogonalization
Language Orthogonalization for Zero-Shot Cross-Lingual Audio Deepfake Detection
Abstract: Audio deepfake detectors need to transfer to languages absent from training, as multilingual speech synthesis outpaces labeled anti-spoofing resources. While detectors increasingly rely on self-supervised speech models (S3Ms), these backbones encode language-dependent structure that confounds spoof cues. We address this confound through language orthogonalization, a target-free ridge map that removes S3M variation projected onto continuous language-identification (LID) embeddings. Across six languages, six S3M backbones, and all Leave-N-Out settings, it consistently reduces EER across unseen languages. Cross-lingual EER correlates with LID-space distance, where orthogonalization yields larger gains for more distant transfers.