Kong Aik Lee

Junyi Peng, Oldřich Plchot, Xiao Song et al.

Modern speaker verification (SV) systems rely on speaker embeddings that are effective but difficult to interpret or query in natural language. Most existing speech-text corpora target controllable synthesis or utterance-level captioning, and provide limited speaker-level supervision for in-the-wild speaker recognition. This paper introduces SpeakerCard-1M, a bilingual speaker-centric resource for evidence-grounded SV, derived from VoxCeleb1/2 and CN-Celeb1/2, where the "-1M" suffix refers to the 1.78M utterance-level captions contained in the release. We adopt a tool-first, LLM-last approach: ten acoustic probes produce field-level evidence, the evidence is aggregated into speaker profiles under a schema that separates relatively stable traits from utterance-level states, and bilingual Speaker Cards are rendered by a constrained LLM that sees only the structured fields. The release includes 56.7K Speaker Card records over 10.2K speakers, 1.78M utterance-level captions, and speaker-ID-disjoint hard-negative triplets. We further define two SV-oriented cross-modal protocols, bidirectional Speaker-Text Retrieval (T2S-R / S2T-R) and Attribute-Conditioned Verification (AC-Verify), and compare a dual-encoder baseline against recent audio language models under a zero-shot forced-choice setting. Joint audio-text training increases VoxCeleb1-O EER by 0.31% absolute over the audio-only baseline. Under a style-symmetric LLM-generated counterfactual protocol, eight recent audio language models (7B-30B+ parameters, both open- and closed-source) score 49-77% on pitch-level AC-Verify under two-way forced choice, compared with 88.66% reached by our dual encoder.

Gaofeng Cheng, Yifan Chen, Runyan Yang et al.

The conversation scenario is one of the most important and most challenging scenarios for speech processing technologies because people in conversation respond to each other in a casual style. Detecting the speech activities of each person in a conversation is vital to downstream tasks, like natural language processing, machine translation, etc. People refer to the detection technology of "who speak when" as speaker diarization (SD). Traditionally, diarization error rate (DER) has been used as the standard evaluation metric of SD systems for a long time. However, DER fails to give enough importance to short conversational phrases, which are short but important on the semantic level. Also, a carefully and accurately manually-annotated testing dataset suitable for evaluating the conversational SD technologies is still unavailable in the speech community. In this paper, we design and describe the Conversational Short-phrases Speaker Diarization (CSSD) task, which consists of training and testing datasets, evaluation metric and baselines. In the dataset aspect, despite the previously open-sourced 180-hour conversational MagicData-RAMC dataset, we prepare an individual 20-hour conversational speech test dataset with carefully and artificially verified speakers timestamps annotations for the CSSD task. In the metric aspect, we design the new conversational DER (CDER) evaluation metric, which calculates the SD accuracy at the utterance level. In the baseline aspect, we adopt a commonly used method: Variational Bayes HMM x-vector system, as the baseline of the CSSD task. Our evaluation metric is publicly available at https://github.com/SpeechClub/CDER_Metric.

Xin Wang, Hector Delgado, Hemlata Tak et al.

ASVspoof 5 is the fifth edition in a series of challenges that promote the study of speech spoofing and deepfake attacks, and the design of detection solutions. Compared to previous challenges, the ASVspoof 5 database is built from crowdsourced data collected from a vastly greater number of speakers in diverse acoustic conditions. Attacks, also crowdsourced, are generated and tested using surrogate detection models, while adversarial attacks are incorporated for the first time. New metrics support the evaluation of spoofing-robust automatic speaker verification (SASV) as well as stand-alone detection solutions, i.e., countermeasures without ASV. We describe the two challenge tracks, the new database, the evaluation metrics, baselines, and the evaluation platform, and present a summary of the results. Attacks significantly compromise the baseline systems, while submissions bring substantial improvements.

Tomi Kinnunen, Kong Aik Lee, Hemlata Tak et al.

Presentation attack (spoofing) detection (PAD) typically operates alongside biometric verification to improve reliablity in the face of spoofing attacks. Even though the two sub-systems operate in tandem to solve the single task of reliable biometric verification, they address different detection tasks and are hence typically evaluated separately. Evidence shows that this approach is suboptimal. We introduce a new metric for the joint evaluation of PAD solutions operating in situ with biometric verification. In contrast to the tandem detection cost function proposed recently, the new tandem equal error rate (t-EER) is parameter free. The combination of two classifiers nonetheless leads to a \emph{set} of operating points at which false alarm and miss rates are equal and also dependent upon the prevalence of attacks. We therefore introduce the \emph{concurrent} t-EER, a unique operating point which is invariable to the prevalence of attacks. Using both modality (and even application) agnostic simulated scores, as well as real scores for a voice biometrics application, we demonstrate application of the t-EER to a wide range of biometric system evaluations under attack. The proposed approach is a strong candidate metric for the tandem evaluation of PAD systems and biometric comparators.

Meng Liu, Kong Aik Lee, Longbiao Wang et al.

Visual speech (i.e., lip motion) is highly related to auditory speech due to the co-occurrence and synchronization in speech production. This paper investigates this correlation and proposes a cross-modal speech co-learning paradigm. The primary motivation of our cross-modal co-learning method is modeling one modality aided by exploiting knowledge from another modality. Specifically, two cross-modal boosters are introduced based on an audio-visual pseudo-siamese structure to learn the modality-transformed correlation. Inside each booster, a max-feature-map embedded Transformer variant is proposed for modality alignment and enhanced feature generation. The network is co-learned both from scratch and with pretrained models. Experimental results on the LRSLip3, GridLip, LomGridLip, and VoxLip datasets demonstrate that our proposed method achieves 60% and 20% average relative performance improvement over independently trained audio-only/visual-only and baseline fusion systems, respectively.

Xiaohui Liu, Meng Liu, Lin Zhang et al.

The Audio Deep Synthesis Detection (ADD) Challenge has been held to detect generated human-like speech. With our submitted system, this paper provides an overall assessment of track 1 (Low-quality Fake Audio Detection) and track 2 (Partially Fake Audio Detection). In this paper, spectro-temporal artifacts were detected using raw temporal signals, spectral features, as well as deep embedding features. To address track 1, low-quality data augmentation, domain adaptation via finetuning, and various complementary feature information fusion were aggregated in our system. Furthermore, we analyzed the clustering characteristics of subsystems with different features by visualization method and explained the effectiveness of our proposed greedy fusion strategy. As for track 2, frame transition and smoothing were detected using self-supervised learning structure to capture the manipulation of PF attacks in the time domain. We ranked 4th and 5th in track 1 and track 2, respectively.

Massimiliano Todisco, Michele Panariello, Xin Wang et al.

We present Malacopula, a neural-based generalised Hammerstein model designed to introduce adversarial perturbations to spoofed speech utterances so that they better deceive automatic speaker verification (ASV) systems. Using non-linear processes to modify speech utterances, Malacopula enhances the effectiveness of spoofing attacks. The model comprises parallel branches of polynomial functions followed by linear time-invariant filters. The adversarial optimisation procedure acts to minimise the cosine distance between speaker embeddings extracted from spoofed and bona fide utterances. Experiments, performed using three recent ASV systems and the ASVspoof 2019 dataset, show that Malacopula increases vulnerabilities by a substantial margin. However, speech quality is reduced and attacks can be detected effectively under controlled conditions. The findings emphasise the need to identify new vulnerabilities and design defences to protect ASV systems from adversarial attacks in the wild.

Kong Aik Lee, Tomi Kinnunen, Daniele Colibro et al.

This manuscript describes the I4U submission to the 2020 NIST Speaker Recognition Evaluation (SRE'20) Conversational Telephone Speech (CTS) Challenge. The I4U's submission was resulted from active collaboration among researchers across eight research teams - I$^2$R (Singapore), UEF (Finland), VALPT (Italy, Spain), NEC (Japan), THUEE (China), LIA (France), NUS (Singapore), INRIA (France) and TJU (China). The submission was based on the fusion of top performing sub-systems and sub-fusion systems contributed by individual teams. Efforts have been spent on the use of common development and validation sets, submission schedule and milestone, minimizing inconsistency in trial list and score file format across sites.

Alexey Sholokhov, Nikita Kuzmin, Kong Aik Lee et al.

This paper focuses on multi-enrollment speaker recognition which naturally occurs in the task of online speaker clustering, and studies the properties of different scoring back-ends in this scenario. First, we show that popular cosine scoring suffers from poor score calibration with a varying number of enrollment utterances. Second, we propose a simple replacement for cosine scoring based on an extremely constrained version of probabilistic linear discriminant analysis (PLDA). The proposed model improves over the cosine scoring for multi-enrollment recognition while keeping the same performance in the case of one-to-one comparisons. Finally, we consider an online speaker clustering task where each step naturally involves multi-enrollment recognition. We propose an online clustering algorithm allowing us to take benefits from the PLDA model such as the ability to handle uncertainty and better score calibration. Our experiments demonstrate the effectiveness of the proposed algorithm.

Tianchi Liu, Kong Aik Lee, Qiongqiong Wang et al.

For speaker recognition, it is difficult to extract an accurate speaker representation from speech because of its mixture of speaker traits and content. This paper proposes a disentanglement framework that simultaneously models speaker traits and content variability in speech. It is realized with the use of three Gaussian inference layers, each consisting of a learnable transition model that extracts distinct speech components. Notably, a strengthened transition model is specifically designed to model complex speech dynamics. We also propose a self-supervision method to dynamically disentangle content without the use of labels other than speaker identities. The efficacy of the proposed framework is validated via experiments conducted on the VoxCeleb and SITW datasets with 9.56% and 8.24% average reductions in EER and minDCF, respectively. Since neither additional model training nor data is specifically needed, it is easily applicable in practical use.

Xin Wang, Héctor Delgado, Nicholas Evans et al.

ASVspoof 5 is the fifth edition in a series of challenges which promote the study of speech spoofing and deepfake detection solutions. A significant change from previous challenge editions is a new crowdsourced database collected from a substantially greater number of speakers under diverse recording conditions, and a mix of cutting-edge and legacy generative speech technology. With the new database described elsewhere, we provide in this paper an overview of the ASVspoof 5 challenge results for the submissions of 53 participating teams. While many solutions perform well, performance degrades under adversarial attacks and the application of neural encoding/compression schemes. Together with a review of post-challenge results, we also report a study of calibration in addition to other principal challenges and outline a road-map for the future of ASVspoof.

Tianchi Liu, Ivan Kukanov, Zihan Pan et al.

The effects of language mismatch impact speech anti-spoofing systems, while investigations and quantification of these effects remain limited. Existing anti-spoofing datasets are mainly in English, and the high cost of acquiring multilingual datasets hinders training language-independent models. We initiate this work by evaluating top-performing speech anti-spoofing systems that are trained on English data but tested on other languages, observing notable performance declines. We propose an innovative approach - Accent-based data expansion via TTS (ACCENT), which introduces diverse linguistic knowledge to monolingual-trained models, improving their cross-lingual capabilities. We conduct experiments on a large-scale dataset consisting of over 3 million samples, including 1.8 million training samples and nearly 1.2 million testing samples across 12 languages. The language mismatch effects are preliminarily quantified and remarkably reduced over 15% by applying the proposed ACCENT. This easily implementable method shows promise for multilingual and low-resource language scenarios.

Nikita Kuzmin, Songting Liu, Kong Aik Lee et al.

Protecting speaker identity is crucial for online voice applications, yet streaming speaker anonymization (SA) remains underexplored. Recent research has demonstrated that neural audio codec (NAC) provides superior speaker feature disentanglement and linguistic fidelity. NAC can also be used with causal language models (LM) to enhance linguistic fidelity and prompt control for streaming tasks. However, existing NAC-based online LM systems are designed for voice conversion (VC) rather than anonymization, lacking the techniques required for privacy protection. Building on these advances, we present Stream-Voice-Anon, which adapts modern causal LM-based NAC architectures specifically for streaming SA by integrating anonymization techniques. Our anonymization approach incorporates pseudo-speaker representation sampling, a speaker embedding mixing and diverse prompt selection strategies for LM conditioning that leverage the disentanglement properties of quantized content codes to prevent speaker information leakage. Additionally, we compare dynamic and fixed delay configurations to explore latency-privacy trade-offs in real-time scenarios. Under the VoicePrivacy 2024 Challenge protocol, Stream-Voice-Anon achieves substantial improvements in intelligibility (up to 46% relative WER reduction) and emotion preservation (up to 28% UAR relative) compared to the previous state-of-the-art streaming method DarkStream while maintaining comparable latency (180ms vs 200ms) and privacy protection against lazy-informed attackers, though showing 15% relative degradation against semi-informed attackers.

Nikita Kuzmin, Kong Aik Lee, Eng Siong Chng

We address the challenge of preserving emotional content in streaming speaker anonymization (SA). Neural audio codec language models trained for audio continuation tend to degrade source emotion: content tokens discard emotional information, and the model defaults to dominant acoustic patterns rather than preserving paralinguistic attributes. We propose supervised finetuning with neutral-emotion utterance pairs from the same speaker, combined with frame-level emotion distillation on acoustic token hidden states. All modifications are confined to finetuning, which takes less than 2 hours on 4 GPUs and adds zero inference latency overhead, while maintaining a competitive 180ms streaming latency. On the VoicePrivacy 2024 protocol, our approach achieves a 49.2% UAR (emotion preservation) with 5.77% WER (intelligibility), a +24% relative UAR improvement over the baseline (39.7%->49.2%) and +10% over the emotion-prompt variant (44.6% UAR), while maintaining strong privacy (EER 49.0%). Demo and code are available: https://anonymous3842031239.github.io/

Jing Yang Lee, Kong Aik Lee, Woon-Seng Gan

To engage human users in meaningful conversation, open-domain dialogue agents are required to generate diverse and contextually coherent dialogue. Despite recent advancements, which can be attributed to the usage of pretrained language models, the generation of diverse and coherent dialogue remains an open research problem. A popular approach to address this issue involves the adaptation of variational frameworks. However, while these approaches successfully improve diversity, they tend to compromise on contextual coherence. Hence, we propose the Bayesian Open-domain Dialogue with Empirical Bayes (BODEB) framework, an empirical bayes framework for constructing an Bayesian open-domain dialogue agent by leveraging pretrained parameters to inform the prior and posterior parameter distributions. Empirical results show that BODEB achieves better results in terms of both diversity and coherence compared to variational frameworks.

Tianchi Liu, Duc-Tuan Truong, Rohan Kumar Das et al.

Speech foundation models have significantly advanced various speech-related tasks by providing exceptional representation capabilities. However, their high-dimensional output features often create a mismatch with downstream task models, which typically require lower-dimensional inputs. A common solution is to apply a dimensionality reduction (DR) layer, but this approach increases parameter overhead, computational costs, and risks losing valuable information. To address these issues, we propose Nested Res2Net (Nes2Net), a lightweight back-end architecture designed to directly process high-dimensional features without DR layers. The nested structure enhances multi-scale feature extraction, improves feature interaction, and preserves high-dimensional information. We first validate Nes2Net on CtrSVDD, a singing voice deepfake detection dataset, and report a 22% performance improvement and an 87% back-end computational cost reduction over the state-of-the-art baseline. Additionally, extensive testing across four diverse datasets: ASVspoof 2021, ASVspoof 5, PartialSpoof, and In-the-Wild, covering fully spoofed speech, adversarial attacks, partial spoofing, and real-world scenarios, consistently highlights Nes2Net's superior robustness and generalization capabilities. The code package and pre-trained models are available at https://github.com/Liu-Tianchi/Nes2Net.

Jinghao He, Zhengyan Sheng, Liping Chen et al.

This paper focuses on explaining the timbre conveyed by speech signals and introduces a task termed voice timbre attribute detection (vTAD). In this task, voice timbre is explained with a set of sensory attributes describing its human perception. A pair of speech utterances is processed, and their intensity is compared in a designated timbre descriptor. Moreover, a framework is proposed, which is built upon the speaker embeddings extracted from the speech utterances. The investigation is conducted on the VCTK-RVA dataset. Experimental examinations on the ECAPA-TDNN and FACodec speaker encoders demonstrated that: 1) the ECAPA-TDNN speaker encoder was more capable in the seen scenario, where the testing speakers were included in the training set; 2) the FACodec speaker encoder was superior in the unseen scenario, where the testing speakers were not part of the training, indicating enhanced generalization capability. The VCTK-RVA dataset and open-source code are available on the website https://github.com/vTAD2025-Challenge/vTAD.

Jing Yang Lee, Kong Aik Lee, Woon-Seng Gan

Despite recent progress in generative open-domain dialogue, the issue of low response diversity persists. Prior works have addressed this issue via either novel objective functions, alternative learning approaches such as variational frameworks, or architectural extensions such as the Randomized Link (RL) Transformer. However, these approaches typically entail either additional difficulties during training/inference, or a significant increase in model size and complexity. Hence, we propose the \underline{Pa}rtially \underline{Ra}ndomized trans\underline{Former} (PaRaFormer), a simple extension of the transformer which involves freezing the weights of selected layers after random initialization. Experimental results reveal that the performance of the PaRaformer is comparable to that of the aforementioned approaches, despite not entailing any additional training difficulty or increase in model complexity.

Junjie Li, Kong Aik Lee

An utterance-level speaker embedding is typically obtained by aggregating a sequence of frame-level representations. However, in real-world scenarios, individual frames encode not only speaker-relevant information but also various nuisance factors. As a result, different frames contribute unequally to the final utterance-level speaker representation for Automatic Speaker Verification systems. To address this issue, we propose to estimate the inherent uncertainty of each frame and assign adaptive weights accordingly, where frames with higher uncertainty receive lower attention. Based on this idea, we present U3-xi, a comprehensive framework designed to produce more reliable and interpretable uncertainty estimates for speaker embeddings. Specifically, we introduce several strategies for uncertainty supervision. First, we propose speaker-level uncertainty supervision via a Stochastic Variance Loss, where the distance between an utterance embedding and its corresponding speaker centroid serves as a pseudo ground truth for uncertainty learning. Second, we incorporate global-level uncertainty supervision by injecting the predicted uncertainty into the sof tmax scale during training. This adaptive scaling mechanism adjusts the sharpness of the decision boundary according to sample difficulty, providing global guidance. Third, we redesign the uncertainty estimation module by integrating a Transformer encoder with multi-view self-attention, enabling the model to capture rich local and long-range temporal dependencies. Comprehensive experiments demonstrate that U3-xi is model-agnostic and can be seamlessly applied to various speaker encoders. In particular, when applied to ECAPA-TDNN, it achieves 21.1% and 15.57% relative improvements on the VoxCeleb1 test sets in terms of EER and minDCF, respectively.

Zhengyan Sheng, Jinghao He, Liping Chen et al.

Voice timbre refers to the unique quality or character of a person's voice that distinguishes it from others as perceived by human hearing. The Voice Timbre Attribute Detection (VtaD) 2025 challenge focuses on explaining the voice timbre attribute in a comparative manner. In this challenge, the human impression of voice timbre is verbalized with a set of sensory descriptors, including bright, coarse, soft, magnetic, and so on. The timbre is explained from the comparison between two voices in their intensity within a specific descriptor dimension. The VtaD 2025 challenge starts in May and culminates in a special proposal at the NCMMSC2025 conference in October 2025 in Zhenjiang, China.

Qiongqiong Wang, Kong Aik Lee

Uncertainty modeling in speaker representation aims to learn the variability present in speech utterances. While the conventional cosine-scoring is computationally efficient and prevalent in speaker recognition, it lacks the capability to handle uncertainty. To address this challenge, this paper proposes an approach for estimating uncertainty at the speaker embedding front-end and propagating it to the cosine scoring back-end. Experiments conducted on the VoxCeleb and SITW datasets confirmed the efficacy of the proposed method in handling uncertainty arising from embedding estimation. It achieved improvement with 8.5% and 9.8% average reductions in EER and minDCF compared to the conventional cosine similarity. It is also computationally efficient in practice.

Weiwei Lin, Chenhang He, Man-Wai Mak et al.

Achieving nuanced and accurate emulation of human voice has been a longstanding goal in artificial intelligence. Although significant progress has been made in recent years, the mainstream of speech synthesis models still relies on supervised speaker modeling and explicit reference utterances. However, there are many aspects of human voice, such as emotion, intonation, and speaking style, for which it is hard to obtain accurate labels. In this paper, we propose VoxGenesis, a novel unsupervised speech synthesis framework that can discover a latent speaker manifold and meaningful voice editing directions without supervision. VoxGenesis is conceptually simple. Instead of mapping speech features to waveforms deterministically, VoxGenesis transforms a Gaussian distribution into speech distributions conditioned and aligned by semantic tokens. This forces the model to learn a speaker distribution disentangled from the semantic content. During the inference, sampling from the Gaussian distribution enables the creation of novel speakers with distinct characteristics. More importantly, the exploration of latent space uncovers human-interpretable directions associated with specific speaker characteristics such as gender attributes, pitch, tone, and emotion, allowing for voice editing by manipulating the latent codes along these identified directions. We conduct extensive experiments to evaluate the proposed VoxGenesis using both subjective and objective metrics, finding that it produces significantly more diverse and realistic speakers with distinct characteristics than the previous approaches. We also show that latent space manipulation produces consistent and human-identifiable effects that are not detrimental to the speech quality, which was not possible with previous approaches. Audio samples of VoxGenesis can be found at: \url{https://bit.ly/VoxGenesis}.

Chenyang Guo, Liping Chen, Zhuhai Li et al.

Neural networks are commonly known to be vulnerable to adversarial attacks mounted through subtle perturbation on the input data. Recent development in voice-privacy protection has shown the positive use cases of the same technique to conceal speaker's voice attribute with additive perturbation signal generated by an adversarial network. This paper examines the reversibility property where an entity generating the adversarial perturbations is authorized to remove them and restore original speech (e.g., the speaker him/herself). A similar technique could also be used by an investigator to deanonymize a voice-protected speech to restore criminals' identities in security and forensic analysis. In this setting, the perturbation generative module is assumed to be known in the removal process. To this end, a joint training of perturbation generation and removal modules is proposed. Experimental results on the LibriSpeech dataset demonstrated that the subtle perturbations added to the original speech can be predicted from the anonymized speech while achieving the goal of privacy protection. By removing these perturbations from the anonymized sample, the original speech can be restored. Audio samples can be found in \url{https://voiceprivacy.github.io/Perturbation-Generation-Removal/}.

Junjie Li, Ke Zhang, Shuai Wang et al.

Audio-visual Target Speaker Extraction (AV-TSE) aims to isolate the speech of a specific target speaker from an audio mixture using time-synchronized visual cues. In real-world scenarios, visual cues are not always available due to various impairments, which undermines the stability of AV-TSE. Despite this challenge, humans can maintain attentional momentum over time, even when the target speaker is not visible. In this paper, we introduce the Momentum Multi-modal target Speaker Extraction (MoMuSE), which retains a speaker identity momentum in memory, enabling the model to continuously track the target speaker. Designed for real-time inference, MoMuSE extracts the current speech window with guidance from both visual cues and dynamically updated speaker momentum. Experimental results demonstrate that MoMuSE exhibits significant improvement, particularly in scenarios with severe impairment of visual cues.

Nikita Kuzmin, Tao Zhong, Jiajun Deng et al.

End-to-end full-duplex speech models feed user audio through an always-on LLM backbone, yet the speaker privacy implications of their hidden representations remain unexamined. Following the VoicePrivacy 2024 protocol with a lazy-informed attacker, we show that the hidden states of SALM-Duplex and Moshi leak substantial speaker identity across all transformer layers. Layer-wise and turn-wise analyses reveal that leakage persists across all layers, with SALM-Duplex showing stronger leakage in early layers while Moshi leaks uniformly, and that Linkability rises sharply within the first few turns. We propose two streaming anonymization setups using Stream-Voice-Anon: a waveform-level front-end (Anon-W2W) and a feature-domain replacement (Anon-W2F). Anon-W2F raises EER by over 3.5x relative to the discrete encoder baseline (11.2% to 41.0%), approaching the 50% random-chance ceiling, while Anon-W2W retains 78-93% of baseline sBERT across setups with sub-second response latency (FRL under 0.8 s).

Haoran Sun, Chen Cai, Huiping Zhuang et al.

The rapid development of deepfake video technology has not only facilitated artistic creation but also made it easier to spread misinformation. Traditional deepfake video detection (DVD) methods face issues such as a lack of transparency in their principles and insufficient generalization capabilities to cope with evolving forgery techniques. This highlights an urgent need for detectors that can identify forged content and provide verifiable reasoning explanations. This paper proposes the explainable deepfake video detection (EDVD) task and designs the EDVD-LLaMA multimodal, a large language model (MLLM) reasoning framework, which provides traceable reasoning processes alongside accurate detection results and trustworthy explanations. Our approach first incorporates a Spatio-Temporal Subtle Information Tokenization (ST-SIT) to extract and fuse global and local cross-frame deepfake features, providing rich spatio-temporal semantic information input for MLLM reasoning. Second, we construct a Fine-grained Multimodal Chain-of-Thought (Fg-MCoT) mechanism, which introduces facial feature data as hard constraints during the reasoning process to achieve pixel-level spatio-temporal video localization, suppress hallucinated outputs, and enhance the reliability of the chain of thought. In addition, we build an Explainable Reasoning FF++ benchmark dataset (ER-FF++set), leveraging structured data to annotate videos and ensure quality control, thereby supporting dual supervision for reasoning and detection. Extensive experiments demonstrate that EDVD-LLaMA achieves outstanding performance and robustness in terms of detection accuracy, explainability, and its ability to handle cross-forgery methods and cross-dataset scenarios. Compared to previous DVD methods, it provides a more explainable and superior solution. The source code and dataset will be publicly available.

Duc-Tuan Truong, Tianchi Liu, Junjie Li et al.

In speech deepfake detection (SDD), data augmentation (DA) is commonly used to improve model generalization across varied speech conditions and spoofing attacks. However, during training, the backpropagated gradients from original and augmented inputs may misalign, which can result in conflicting parameter updates. These conflicts could hinder convergence and push the model toward suboptimal solutions, thereby reducing the benefits of DA. To investigate and address this issue, we design a dual-path data-augmented (DPDA) training framework with gradient alignment for SDD. In our framework, each training utterance is processed through two input paths: one using the original speech and the other with its augmented version. This design allows us to compare and align their backpropagated gradient directions to reduce optimization conflicts. Our analysis shows that approximately 25% of training iterations exhibit gradient conflicts between the original inputs and their augmented counterparts when using RawBoost augmentation. By resolving these conflicts with gradient alignment, our method accelerates convergence by reducing the number of training epochs and achieves up to an 18.69% relative reduction in Equal Error Rate on the In-the-Wild dataset compared to the baseline.

Duc-Tuan Truong, Tianchi Liu, Ruijie Tao et al.

Recent work shows that one-class learning can detect unseen deepfake attacks by modeling a compact distribution of bona fide speech around a single centroid. However, the single-centroid assumption can oversimplify the bona fide speech representation and overlook useful cues, such as speech quality, which reflects the naturalness of the speech. Speech quality can be easily obtained using existing speech quality assessment models that estimate it through Mean Opinion Score. In this paper, we propose QAMO: Quality-Aware Multi-Centroid One-Class Learning for speech deepfake detection. QAMO extends conventional one-class learning by introducing multiple quality-aware centroids. In QAMO, each centroid is optimized to represent a distinct speech quality subspaces, enabling better modeling of intra-class variability in bona fide speech. In addition, QAMO supports a multi-centroid ensemble scoring strategy, which improves decision thresholding and reduces the need for quality labels during inference. With two centroids to represent high- and low-quality speech, our proposed QAMO achieves an equal error rate of 5.09% in In-the-Wild dataset, outperforming previous one-class and quality-aware systems.

Liping Chen, Jinghao He, Zhengyan Sheng et al.

The first voice timbre attribute detection challenge is featured in a special session at NCMMSC 2025. It focuses on the explainability of voice timbre and compares the intensity of two speech utterances in a specified timbre descriptor dimension. The evaluation was conducted on the VCTK-RVA dataset. Participants developed their systems and submitted their outputs to the organizer, who evaluated the performance and sent feedback to them. Six teams submitted their outputs, with five providing descriptions of their methodologies.

Duc-Tuan Truong, Ruijie Tao, Tuan Nguyen et al.

Recent synthetic speech detectors leveraging the Transformer model have superior performance compared to the convolutional neural network counterparts. This improvement could be due to the powerful modeling ability of the multi-head self-attention (MHSA) in the Transformer model, which learns the temporal relationship of each input token. However, artifacts of synthetic speech can be located in specific regions of both frequency channels and temporal segments, while MHSA neglects this temporal-channel dependency of the input sequence. In this work, we proposed a Temporal-Channel Modeling (TCM) module to enhance MHSA's capability for capturing temporal-channel dependencies. Experimental results on the ASVspoof 2021 show that with only 0.03M additional parameters, the TCM module can outperform the state-of-the-art system by 9.25% in EER. Further ablation study reveals that utilizing both temporal and channel information yields the most improvement for detecting synthetic speech.

Rui Wang, Liping Chen, Kong AiK Lee et al.

Voice anonymization has been developed as a technique for preserving privacy by replacing the speaker's voice in a speech signal with that of a pseudo-speaker, thereby obscuring the original voice attributes from machine recognition and human perception. In this paper, we focus on altering the voice attributes against machine recognition while retaining human perception. We referred to this as the asynchronous voice anonymization. To this end, a speech generation framework incorporating a speaker disentanglement mechanism is employed to generate the anonymized speech. The speaker attributes are altered through adversarial perturbation applied on the speaker embedding, while human perception is preserved by controlling the intensity of perturbation. Experiments conducted on the LibriSpeech dataset showed that the speaker attributes were obscured with their human perception preserved for 60.71% of the processed utterances.

Xuechen Liu, Md Sahidullah, Kong Aik Lee et al.

It is now well-known that automatic speaker verification (ASV) systems can be spoofed using various types of adversaries. The usual approach to counteract ASV systems against such attacks is to develop a separate spoofing countermeasure (CM) module to classify speech input either as a bonafide, or a spoofed utterance. Nevertheless, such a design requires additional computation and utilization efforts at the authentication stage. An alternative strategy involves a single monolithic ASV system designed to handle both zero-effort imposter (non-targets) and spoofing attacks. Such spoof-aware ASV systems have the potential to provide stronger protections and more economic computations. To this end, we propose to generalize the standalone ASV (G-SASV) against spoofing attacks, where we leverage limited training data from CM to enhance a simple backend in the embedding space, without the involvement of a separate CM module during the test (authentication) phase. We propose a novel yet simple backend classifier based on deep neural networks and conduct the study via domain adaptation and multi-task integration of spoof embeddings at the training stage. Experiments are conducted on the ASVspoof 2019 logical access dataset, where we improve the performance of statistical ASV backends on the joint (bonafide and spoofed) and spoofed conditions by a maximum of 36.2% and 49.8% in terms of equal error rates, respectively.

Sung Hwan Mun, Hye-jin Shim, Hemlata Tak et al.

This study aims to develop a single integrated spoofing-aware speaker verification (SASV) embeddings that satisfy two aspects. First, rejecting non-target speakers' input as well as target speakers' spoofed inputs should be addressed. Second, competitive performance should be demonstrated compared to the fusion of automatic speaker verification (ASV) and countermeasure (CM) embeddings, which outperformed single embedding solutions by a large margin in the SASV2022 challenge. We analyze that the inferior performance of single SASV embeddings comes from insufficient amount of training data and distinct nature of ASV and CM tasks. To this end, we propose a novel framework that includes multi-stage training and a combination of loss functions. Copy synthesis, combined with several vocoders, is also exploited to address the lack of spoofed data. Experimental results show dramatic improvements, achieving a SASV-EER of 1.06% on the evaluation protocol of the SASV2022 challenge.

Jing Yang Lee, Kong Aik Lee, Woon Seng Gan

In recent years, latent variable models, such as the Conditional Variational Auto Encoder (CVAE), have been applied to both personalized and empathetic dialogue generation. Prior work have largely focused on generating diverse dialogue responses that exhibit persona consistency and empathy. However, when it comes to the contextual coherence of the generated responses, there is still room for improvement. Hence, to improve the contextual coherence, we propose a novel Uncertainty Aware CVAE (UA-CVAE) framework. The UA-CVAE framework involves approximating and incorporating the aleatoric uncertainty during response generation. We apply our framework to both personalized and empathetic dialogue generation. Empirical results show that our framework significantly improves the contextual coherence of the generated response. Additionally, we introduce a novel automatic metric for measuring contextual coherence, which was found to correlate positively with human judgement.

Fan Yu, Shiliang Zhang, Pengcheng Guo et al.

The ICASSP 2022 Multi-channel Multi-party Meeting Transcription Grand Challenge (M2MeT) focuses on one of the most valuable and the most challenging scenarios of speech technologies. The M2MeT challenge has particularly set up two tracks, speaker diarization (track 1) and multi-speaker automatic speech recognition (ASR) (track 2). Along with the challenge, we released 120 hours of real-recorded Mandarin meeting speech data with manual annotation, including far-field data collected by 8-channel microphone array as well as near-field data collected by each participants' headset microphone. We briefly describe the released dataset, track setups, baselines and summarize the challenge results and major techniques used in the submissions.

Tianchi Liu, Rohan Kumar Das, Kong Aik Lee et al.

The time delay neural network (TDNN) represents one of the state-of-the-art of neural solutions to text-independent speaker verification. However, they require a large number of filters to capture the speaker characteristics at any local frequency region. In addition, the performance of such systems may degrade under short utterance scenarios. To address these issues, we propose a multi-scale frequency-channel attention (MFA), where we characterize speakers at different scales through a novel dual-path design which consists of a convolutional neural network and TDNN. We evaluate the proposed MFA on the VoxCeleb database and observe that the proposed framework with MFA can achieve state-of-the-art performance while reducing parameters and computation complexity. Further, the MFA mechanism is found to be effective for speaker verification with short test utterances.

Jing Yang Lee, Kong Aik Lee, Woon Seng Gan

The generation of personalized dialogue is vital to natural and human-like conversation. Typically, personalized dialogue generation models involve conditioning the generated response on the dialogue history and a representation of the persona/personality of the interlocutor. As it is impractical to obtain the persona/personality representations for every interlocutor, recent works have explored the possibility of generating personalized dialogue by finetuning the model with dialogue examples corresponding to a given persona instead. However, in real-world implementations, a sufficient number of corresponding dialogue examples are also rarely available. Hence, in this paper, we propose a Dual Latent Variable Generator (DLVGen) capable of generating personalized dialogue in the absence of any persona/personality information or any corresponding dialogue examples. Unlike prior work, DLVGen models the latent distribution over potential responses as well as the latent distribution over the agent's potential persona. During inference, latent variables are sampled from both distributions and fed into the decoder. Empirical results show that DLVGen is capable of generating diverse responses which accurately incorporate the agent's persona.

Yi Ma, Kong Aik Lee, Ville Hautamaki et al.

Speech enhancement aims to improve the perceptual quality of the speech signal by suppression of the background noise. However, excessive suppression may lead to speech distortion and speaker information loss, which degrades the performance of speaker embedding extraction. To alleviate this problem, we propose an end-to-end deep learning framework, dubbed PL-EESR, for robust speaker representation extraction. This framework is optimized based on the feedback of the speaker identification task and the high-level perceptual deviation between the raw speech signal and its noisy version. We conducted speaker verification tasks in both noisy and clean environment respectively to evaluate our system. Compared to the baseline, our method shows better performance in both clean and noisy environments, which means our method can not only enhance the speaker relative information but also avoid adding distortions.

Junichi Yamagishi, Xin Wang, Massimiliano Todisco et al.

ASVspoof 2021 is the forth edition in the series of bi-annual challenges which aim to promote the study of spoofing and the design of countermeasures to protect automatic speaker verification systems from manipulation. In addition to a continued focus upon logical and physical access tasks in which there are a number of advances compared to previous editions, ASVspoof 2021 introduces a new task involving deepfake speech detection. This paper describes all three tasks, the new databases for each of them, the evaluation metrics, four challenge baselines, the evaluation platform and a summary of challenge results. Despite the introduction of channel and compression variability which compound the difficulty, results for the logical access and deepfake tasks are close to those from previous ASVspoof editions. Results for the physical access task show the difficulty in detecting attacks in real, variable physical spaces. With ASVspoof 2021 being the first edition for which participants were not provided with any matched training or development data and with this reflecting real conditions in which the nature of spoofed and deepfake speech can never be predicated with confidence, the results are extremely encouraging and demonstrate the substantial progress made in the field in recent years.

Héctor Delgado, Nicholas Evans, Tomi Kinnunen et al.

The automatic speaker verification spoofing and countermeasures (ASVspoof) challenge series is a community-led initiative which aims to promote the consideration of spoofing and the development of countermeasures. ASVspoof 2021 is the 4th in a series of bi-annual, competitive challenges where the goal is to develop countermeasures capable of discriminating between bona fide and spoofed or deepfake speech. This document provides a technical description of the ASVspoof 2021 challenge, including details of training, development and evaluation data, metrics, baselines, evaluation rules, submission procedures and the schedule.

Jean-Francois Bonastre, Hector Delgado, Nicholas Evans et al.

For many decades, research in speech technologies has focused upon improving reliability. With this now meeting user expectations for a range of diverse applications, speech technology is today omni-present. As result, a focus on security and privacy has now come to the fore. Here, the research effort is in its relative infancy and progress calls for greater, multidisciplinary collaboration with security, privacy, legal and ethical experts among others. Such collaboration is now underway. To help catalyse the efforts, this paper provides a high-level overview of some related research. It targets the non-speech audience and describes the benchmarking methodology that has spearheaded progress in traditional research and which now drives recent security and privacy initiatives related to voice biometrics. We describe: the ASVspoof challenge relating to the development of spoofing countermeasures; the VoicePrivacy initiative which promotes research in anonymisation for privacy preservation.

Qiongqiong Wang, Kong Aik Lee, Takafumi Koshinaka et al.

This paper proposes the use of two task-aware warping factors in mask-based speech enhancement (SE). One controls the balance between speech-maintenance and noise-removal in training phases, while the other controls SE power applied to specific downstream tasks in testing phases. Our intention is to alleviate the problem that SE systems trained to improve speech quality often fail to improve other downstream tasks, such as automatic speaker verification (ASV) and automatic speech recognition (ASR), because they do not share the same objects. It is easy to apply the proposed dual-warping factors approach to any mask-based SE method, and it allows a single SE system to handle multiple tasks without task-dependent training. The effectiveness of our proposed approach has been confirmed on the SITW dataset for ASV evaluation and the LibriSpeech dataset for ASR and speech quality evaluations of 0-20dB. We show that different warping values are necessary for a single SE to achieve optimal performance w.r.t. the three tasks. With the use of task-dependent warping factors, speech quality was improved by an 84.7% PESQ increase, ASV had a 22.4% EER reduction, and ASR had a 52.2% WER reduction, on 0dB speech. The effectiveness of the task-dependent warping factors were also cross-validated on VoxCeleb-1 test set for ASV and LibriSpeech dev-clean set for ASV and quality evaluations. The proposed method is highly effective and easy to apply in practice.

Kong Aik Lee, Qiongqiong Wang, Takafumi Koshinaka

We present a Bayesian formulation for deep speaker embedding, wherein the xi-vector is the Bayesian counterpart of the x-vector, taking into account the uncertainty estimate. On the technology front, we offer a simple and straightforward extension to the now widely used x-vector. It consists of an auxiliary neural net predicting the frame-wise uncertainty of the input sequence. We show that the proposed extension leads to substantial improvement across all operating points, with a significant reduction in error rates and detection cost. On the theoretical front, our proposal integrates the Bayesian formulation of linear Gaussian model to speaker-embedding neural networks via the pooling layer. In one sense, our proposal integrates the Bayesian formulation of the i-vector to that of the x-vector. Hence, we refer to the embedding as the xi-vector, which is pronounced as /zai/ vector. Experimental results on the SITW evaluation set show a consistent improvement of over 17.5% in equal-error-rate and 10.9% in minimum detection cost.

Jing Yang Lee, Kong Aik Lee, Woon Seng Gan

Conventional approaches to personalized dialogue generation typically require a large corpus, as well as predefined persona information. However, in a real-world setting, neither a large corpus of training data nor persona information are readily available. To address these practical limitations, we propose a novel multi-task meta-learning approach which involves training a model to adapt to new personas without relying on a large corpus, or on any predefined persona information. Instead, the model is tasked with generating personalized responses based on only the dialogue context. Unlike prior work, our approach leverages on the provided persona information only during training via the introduction of an auxiliary persona reconstruction task. In this paper, we introduce 2 frameworks that adopt the proposed multi-task meta-learning approach: the Multi-Task Meta-Learning (MTML) framework, and the Alternating Multi-Task Meta-Learning (AMTML) framework. Experimental results show that utilizing MTML and AMTML results in dialogue responses with greater persona consistency.

Hongning Zhu, Kong Aik Lee, Haizhou Li

This paper proposes a serialized multi-layer multi-head attention for neural speaker embedding in text-independent speaker verification. In prior works, frame-level features from one layer are aggregated to form an utterance-level representation. Inspired by the Transformer network, our proposed method utilizes the hierarchical architecture of stacked self-attention mechanisms to derive refined features that are more correlated with speakers. Serialized attention mechanism contains a stack of self-attention modules to create fixed-dimensional representations of speakers. Instead of utilizing multi-head attention in parallel, the proposed serialized multi-layer multi-head attention is designed to aggregate and propagate attentive statistics from one layer to the next in a serialized manner. In addition, we employ an input-aware query for each utterance with the statistics pooling. With more layers stacked, the neural network can learn more discriminative speaker embeddings. Experiment results on VoxCeleb1 dataset and SITW dataset show that our proposed method outperforms other baseline methods, including x-vectors and other x-vectors + conventional attentive pooling approaches by 9.7% in EER and 8.1% in DCF0.01.

Li Zhang, Qing Wang, Kong Aik Lee et al.

In far-field speaker verification, the performance of speaker embeddings is susceptible to degradation when there is a mismatch between the conditions of enrollment and test speech. To solve this problem, we propose the feature-level and instance-level transfer learning in the teacher-student framework to learn a domain-invariant embedding space. For the feature-level knowledge transfer, we develop the contrastive loss to transfer knowledge from teacher model to student model, which can not only decrease the intra-class distance, but also enlarge the inter-class distance. Moreover, we propose the instance-level pairwise distance transfer method to force the student model to preserve pairwise instances distance from the well optimized embedding space of the teacher model. On FFSVC 2020 evaluation set, our EER on Full-eval trials is relatively reduced by 13.9% compared with the fusion system result on Partial-eval trials of Task2. On Task1, compared with the winner's DenseNet result on Partial-eval trials, our minDCF on Full-eval trials is relatively reduced by 6.3%. On Task3, the EER and minDCF of our proposed method on Full-eval trials are very close to the result of the fusion system on Partial-eval trials. Our results also outperform other competitive domain adaptation methods.

Tomi Kinnunen, Andreas Nautsch, Md Sahidullah et al.

Whether it be for results summarization, or the analysis of classifier fusion, some means to compare different classifiers can often provide illuminating insight into their behaviour, (dis)similarity or complementarity. We propose a simple method to derive 2D representation from detection scores produced by an arbitrary set of binary classifiers in response to a common dataset. Based upon rank correlations, our method facilitates a visual comparison of classifiers with arbitrary scores and with close relation to receiver operating characteristic (ROC) and detection error trade-off (DET) analyses. While the approach is fully versatile and can be applied to any detection task, we demonstrate the method using scores produced by automatic speaker verification and voice anti-spoofing systems. The former are produced by a Gaussian mixture model system trained with VoxCeleb data whereas the latter stem from submissions to the ASVspoof 2019 challenge.

Meng Liu, Longbiao Wang, Kong Aik Lee et al.

Audio-visual (AV) lip biometrics is a promising authentication technique that leverages the benefits of both the audio and visual modalities in speech communication. Previous works have demonstrated the usefulness of AV lip biometrics. However, the lack of a sizeable AV database hinders the exploration of deep-learning-based audio-visual lip biometrics. To address this problem, we compile a moderate-size database using existing public databases. Meanwhile, we establish the DeepLip AV lip biometrics system realized with a convolutional neural network (CNN) based video module, a time-delay neural network (TDNN) based audio module, and a multimodal fusion module. Our experiments show that DeepLip outperforms traditional speaker recognition models in context modeling and achieves over 50% relative improvements compared with our best single modality baseline, with an equal error rate of 0.75% and 1.11% on the test datasets, respectively.

Andreas Nautsch, Xin Wang, Nicholas Evans et al.

The ASVspoof initiative was conceived to spearhead research in anti-spoofing for automatic speaker verification (ASV). This paper describes the third in a series of bi-annual challenges: ASVspoof 2019. With the challenge database and protocols being described elsewhere, the focus of this paper is on results and the top performing single and ensemble system submissions from 62 teams, all of which out-perform the two baseline systems, often by a substantial margin. Deeper analyses shows that performance is dominated by specific conditions involving either specific spoofing attacks or specific acoustic environments. While fusion is shown to be particularly effective for the logical access scenario involving speech synthesis and voice conversion attacks, participants largely struggled to apply fusion successfully for the physical access scenario involving simulated replay attacks. This is likely the result of a lack of system complementarity, while oracle fusion experiments show clear potential to improve performance. Furthermore, while results for simulated data are promising, experiments with real replay data show a substantial gap, most likely due to the presence of additive noise in the latter. This finding, among others, leads to a number of ideas for further research and directions for future editions of the ASVspoof challenge.

Qiongqiong Wang, Kong Aik Lee, Takafumi Koshinaka

This paper presents a simple but effective method that uses multi-resolution feature maps with convolutional neural networks (CNNs) for anti-spoofing in automatic speaker verification (ASV). The central idea is to alleviate the problem that the feature maps commonly used in anti-spoofing networks are insufficient for building discriminative representations of audio segments, as they are often extracted by a single-length sliding window. Resulting trade-offs between time and frequency resolutions restrict the information in single spectrograms. The proposed method improves both frequency resolution and time resolution by stacking multiple spectrograms that are extracted using different window lengths. These are fed into a convolutional neural network in the form of multiple channels, making it possible to extract more information from input signals while only marginally increasing computational costs. The efficiency of the proposed method has been conformed on the ASVspoof 2019 database. We show that the use of the proposed multiresolution inputs consistently outperforms that of score fusion across different CNN architectures. Moreover, computational cost remains small.