Haibin Wu

Abdelrahman Abouelenin, Atabak Ashfaq, Adam Atkinson et al. · microsoft-research

We introduce Phi-4-Mini and Phi-4-Multimodal, compact yet highly capable language and multimodal models. Phi-4-Mini is a 3.8-billion-parameter language model trained on high-quality web and synthetic data, significantly outperforming recent open-source models of similar size and matching the performance of models twice its size on math and coding tasks requiring complex reasoning. This achievement is driven by a carefully curated synthetic data recipe emphasizing high-quality math and coding datasets. Compared to its predecessor, Phi-3.5-Mini, Phi-4-Mini features an expanded vocabulary size of 200K tokens to better support multilingual applications, as well as group query attention for more efficient long-sequence generation. Phi-4-Multimodal is a multimodal model that integrates text, vision, and speech/audio input modalities into a single model. Its novel modality extension approach leverages LoRA adapters and modality-specific routers to allow multiple inference modes combining various modalities without interference. For example, it now ranks first in the OpenASR leaderboard to date, although the LoRA component of the speech/audio modality has just 460 million parameters. Phi-4-Multimodal supports scenarios involving (vision + language), (vision + speech), and (speech/audio) inputs, outperforming larger vision-language and speech-language models on a wide range of tasks. Additionally, we experiment to further train Phi-4-Mini to enhance its reasoning capabilities. Despite its compact 3.8-billion-parameter size, this experimental version achieves reasoning performance on par with or surpassing significantly larger models, including DeepSeek-R1-Distill-Qwen-7B and DeepSeek-R1-Distill-Llama-8B.

Chien-yu Huang, Ke-Han Lu, Shih-Heng Wang et al. · cmu, meta-ai

Text language models have shown remarkable zero-shot capability in generalizing to unseen tasks when provided with well-formulated instructions. However, existing studies in speech processing primarily focus on limited or specific tasks. Moreover, the lack of standardized benchmarks hinders a fair comparison across different approaches. Thus, we present Dynamic-SUPERB, a benchmark designed for building universal speech models capable of leveraging instruction tuning to perform multiple tasks in a zero-shot fashion. To achieve comprehensive coverage of diverse speech tasks and harness instruction tuning, we invite the community to collaborate and contribute, facilitating the dynamic growth of the benchmark. To initiate, Dynamic-SUPERB features 55 evaluation instances by combining 33 tasks and 22 datasets. This spans a broad spectrum of dimensions, providing a comprehensive platform for evaluation. Additionally, we propose several approaches to establish benchmark baselines. These include the utilization of speech models, text language models, and the multimodal encoder. Evaluation results indicate that while these baselines perform reasonably on seen tasks, they struggle with unseen ones. We release all materials to the public and welcome researchers to collaborate on the project, advancing technologies in the field together.

Tzu-hsun Feng, Annie Dong, Ching-Feng Yeh et al. · meta-ai, mit

We present the SUPERB challenge at SLT 2022, which aims at learning self-supervised speech representation for better performance, generalization, and efficiency. The challenge builds upon the SUPERB benchmark and implements metrics to measure the computation requirements of self-supervised learning (SSL) representation and to evaluate its generalizability and performance across the diverse SUPERB tasks. The SUPERB benchmark provides comprehensive coverage of popular speech processing tasks, from speech and speaker recognition to audio generation and semantic understanding. As SSL has gained interest in the speech community and showed promising outcomes, we envision the challenge to uplevel the impact of SSL techniques by motivating more practical designs of techniques beyond task performance. We summarize the results of 14 submitted models in this paper. We also discuss the main findings from those submissions and the future directions of SSL research.

Yuan Tseng, Layne Berry, Yi-Ting Chen et al. · meta-ai, mit

Audio-visual representation learning aims to develop systems with human-like perception by utilizing correlation between auditory and visual information. However, current models often focus on a limited set of tasks, and generalization abilities of learned representations are unclear. To this end, we propose the AV-SUPERB benchmark that enables general-purpose evaluation of unimodal audio/visual and bimodal fusion representations on 7 datasets covering 5 audio-visual tasks in speech and audio processing. We evaluate 5 recent self-supervised models and show that none of these models generalize to all tasks, emphasizing the need for future study on improving universal model performance. In addition, we show that representations may be improved with intermediate-task fine-tuning and audio event classification with AudioSet serves as a strong intermediate task. We release our benchmark with evaluation code and a model submission platform to encourage further research in audio-visual learning.

Kai-Wei Chang, Haibin Wu, Yu-Kai Wang et al. · meta-ai, mit

Prompting has become a practical method for utilizing pre-trained language models (LMs). This approach offers several advantages. It allows an LM to adapt to new tasks with minimal training and parameter updates, thus achieving efficiency in both storage and computation. Additionally, prompting modifies only the LM's inputs and harnesses the generative capabilities of language models to address various downstream tasks in a unified manner. This significantly reduces the need for human labor in designing task-specific models. These advantages become even more evident as the number of tasks served by the LM scales up. Motivated by the strengths of prompting, we are the first to explore the potential of prompting speech LMs in the domain of speech processing. Recently, there has been a growing interest in converting speech into discrete units for language modeling. Our pioneer research demonstrates that these quantized speech units are highly versatile within our unified prompting framework. Not only can they serve as class labels, but they also contain rich phonetic information that can be re-synthesized back into speech signals for speech generation tasks. Specifically, we reformulate speech processing tasks into speech-to-unit generation tasks. As a result, we can seamlessly integrate tasks such as speech classification, sequence generation, and speech generation within a single, unified prompting framework. The experiment results show that the prompting method can achieve competitive performance compared to the strong fine-tuning method based on self-supervised learning models with a similar number of trainable parameters. The prompting method also shows promising results in the few-shot setting. Moreover, with the advanced speech LMs coming into the stage, the proposed prompting framework attains great potential.

Haibin Wu, Kai-Wei Chang, Yuan-Kuei Wu et al.

Large language models (LLMs) have gained considerable attention for Artificial Intelligence Generated Content (AIGC), particularly with the emergence of ChatGPT. However, the direct adaptation of continuous speech to LLMs that process discrete tokens remains an unsolved challenge, hindering the application of LLMs for speech generation. The advanced speech LMs are in the corner, as that speech signals encapsulate a wealth of information, including speaker and emotion, beyond textual data alone. Prompt tuning has demonstrated notable gains in parameter efficiency and competitive performance on some speech classification tasks. However, the extent to which prompts can effectively elicit generation tasks from speech LMs remains an open question. In this paper, we present pioneering research that explores the application of prompt tuning to stimulate speech LMs for various generation tasks, within a unified framework called SpeechGen, with around 10M trainable parameters. The proposed unified framework holds great promise for efficiency and effectiveness, particularly with the imminent arrival of advanced speech LMs, which will significantly enhance the capabilities of the framework. The code and demos of SpeechGen will be available on the project website: \url{https://ga642381.github.io/SpeechPrompt/speechgen}

Quanyi Li, Zhenghao Peng, Haibin Wu et al.

Human-AI shared control allows human to interact and collaborate with AI to accomplish control tasks in complex environments. Previous Reinforcement Learning (RL) methods attempt the goal-conditioned design to achieve human-controllable policies at the cost of redesigning the reward function and training paradigm. Inspired by the neuroscience approach to investigate the motor cortex in primates, we develop a simple yet effective frequency-based approach called \textit{Policy Dissection} to align the intermediate representation of the learned neural controller with the kinematic attributes of the agent behavior. Without modifying the neural controller or retraining the model, the proposed approach can convert a given RL-trained policy into a human-interactive policy. We evaluate the proposed approach on the RL tasks of autonomous driving and locomotion. The experiments show that human-AI shared control achieved by Policy Dissection in driving task can substantially improve the performance and safety in unseen traffic scenes. With human in the loop, the locomotion robots also exhibit versatile controllable motion skills even though they are only trained to move forward. Our results suggest the promising direction of implementing human-AI shared autonomy through interpreting the learned representation of the autonomous agents. Demo video and code will be made available at https://metadriverse.github.io/policydissect.

Haibin Wu, Lingwei Meng, Jiawen Kang et al.

Recently, many novel techniques have been introduced to deal with spoofing attacks, and achieve promising countermeasure (CM) performances. However, these works only take the stand-alone CM models into account. Nowadays, a spoofing aware speaker verification (SASV) challenge which aims to facilitate the research of integrated CM and ASV models, arguing that jointly optimizing CM and ASV models will lead to better performance, is taking place. In this paper, we propose a novel multi-model and multi-level fusion strategy to tackle the SASV task. Compared with purely scoring fusion and embedding fusion methods, this framework first utilizes embeddings from CM models, propagating CM embeddings into a CM block to obtain a CM score. In the second-level fusion, the CM score and ASV scores directly from ASV systems will be concatenated into a prediction block for the final decision. As a result, the best single fusion system has achieved the SASV-EER of 0.97% on the evaluation set. Then by ensembling the top-5 fusion systems, the final SASV-EER reached 0.89%.

Xuanjun Chen, Haibin Wu, Chung-Che Wang et al.

Audio-visual synchronization aims to determine whether the mouth movements and speech in the video are synchronized. VocaLiST reaches state-of-the-art performance by incorporating multimodal Transformers to model audio-visual interact information. However, it requires high computing resources, making it impractical for real-world applications. This paper proposed an MTDVocaLiST model, which is trained by our proposed multimodal Transformer distillation (MTD) loss. MTD loss enables MTDVocaLiST model to deeply mimic the cross-attention distribution and value-relation in the Transformer of VocaLiST. Additionally, we harness uncertainty weighting to fully exploit the interaction information across all layers. Our proposed method is effective in two aspects: From the distillation method perspective, MTD loss outperforms other strong distillation baselines. From the distilled model's performance perspective: 1) MTDVocaLiST outperforms similar-size SOTA models, SyncNet, and Perfect Match models by 15.65% and 3.35%; 2) MTDVocaLiST reduces the model size of VocaLiST by 83.52%, yet still maintaining similar performance.

Haibin Wu, Jiawen Kang, Lingwei Meng et al.

Recent years have witnessed the extraordinary development of automatic speaker verification (ASV). However, previous works show that state-of-the-art ASV models are seriously vulnerable to voice spoofing attacks, and the recently proposed high-performance spoofing countermeasure (CM) models only focus solely on the standalone anti-spoofing tasks, and ignore the subsequent speaker verification process. How to integrate the CM and ASV together remains an open question. A spoofing aware speaker verification (SASV) challenge has recently taken place with the argument that better performance can be delivered when both CM and ASV subsystems are optimized jointly. Under the challenge's scenario, the integrated systems proposed by the participants are required to reject both impostor speakers and spoofing attacks from target speakers, which intuitively and effectively matches the expectation of a reliable, spoofing-robust ASV system. This work focuses on fusion-based SASV solutions and proposes a multi-model fusion framework to leverage the power of multiple state-of-the-art ASV and CM models. The proposed framework vastly improves the SASV-EER from 8.75% to 1.17\%, which is 86% relative improvement compared to the best baseline system in the SASV challenge.

Haibin Wu, Xiaofei Wang, Sefik Emre Eskimez et al.

People change their tones of voice, often accompanied by nonverbal vocalizations (NVs) such as laughter and cries, to convey rich emotions. However, most text-to-speech (TTS) systems lack the capability to generate speech with rich emotions, including NVs. This paper introduces EmoCtrl-TTS, an emotion-controllable zero-shot TTS that can generate highly emotional speech with NVs for any speaker. EmoCtrl-TTS leverages arousal and valence values, as well as laughter embeddings, to condition the flow-matching-based zero-shot TTS. To achieve high-quality emotional speech generation, EmoCtrl-TTS is trained using more than 27,000 hours of expressive data curated based on pseudo-labeling. Comprehensive evaluations demonstrate that EmoCtrl-TTS excels in mimicking the emotions of audio prompts in speech-to-speech translation scenarios. We also show that EmoCtrl-TTS can capture emotion changes, express strong emotions, and generate various NVs in zero-shot TTS. See https://aka.ms/emoctrl-tts for demo samples.

Xuanjun Chen, Haibin Wu, Helen Meng et al.

Audio-visual active speaker detection (AVASD) is well-developed, and now is an indispensable front-end for several multi-modal applications. However, to the best of our knowledge, the adversarial robustness of AVASD models hasn't been investigated, not to mention the effective defense against such attacks. In this paper, we are the first to reveal the vulnerability of AVASD models under audio-only, visual-only, and audio-visual adversarial attacks through extensive experiments. What's more, we also propose a novel audio-visual interaction loss (AVIL) for making attackers difficult to find feasible adversarial examples under an allocated attack budget. The loss aims at pushing the inter-class embeddings to be dispersed, namely non-speech and speech clusters, sufficiently disentangled, and pulling the intra-class embeddings as close as possible to keep them compact. Experimental results show the AVIL outperforms the adversarial training by 33.14 mAP (%) under multi-modal attacks.

Haibin Wu, Ke Tan, Buye Xu et al.

Despite multiple efforts made towards adopting complex-valued deep neural networks (DNNs), it remains an open question whether complex-valued DNNs are generally more effective than real-valued DNNs for monaural speech enhancement. This work is devoted to presenting a critical assessment by systematically examining complex-valued DNNs against their real-valued counterparts. Specifically, we investigate complex-valued DNN atomic units, including linear layers, convolutional layers, long short-term memory (LSTM), and gated linear units. By comparing complex- and real-valued versions of fundamental building blocks in the recently developed gated convolutional recurrent network (GCRN), we show how different mechanisms for basic blocks affect the performance. We also find that the use of complex-valued operations hinders the model capacity when the model size is small. In addition, we examine two recent complex-valued DNNs, i.e. deep complex convolutional recurrent network (DCCRN) and deep complex U-Net (DCUNET). Evaluation results show that both DNNs produce identical performance to their real-valued counterparts while requiring much more computation. Based on these comprehensive comparisons, we conclude that complex-valued DNNs do not provide a performance gain over their real-valued counterparts for monaural speech enhancement, and thus are less desirable due to their higher computational costs.

Leying Zhang, Bowen Shi, Haibin Wu et al.

The rapid advancement of generative audio models has outpaced the development of robust evaluation methodologies. Existing objective metrics and general multimodal large language models (MLLMs) often struggle with domain generalization, zero-shot capabilities, and instructional flexibility. To address these bottlenecks, we propose JASTIN, a generalizable, instruction-driven audio evaluation framework that formulates audio assessment as a self-instructed reasoning task. JASTIN bridges a frozen high-performance audio encoder with a fine-tuned LLM backbone via a trainable audio adapter. To ensure robust zero-shot generalization, we introduce a comprehensive instruction following data preparation pipeline, incorporating Multi-Source, Multi-Task, Multi-Calibration, and Multi-Description data. Experimental results demonstrate that JASTIN achieves state-of-the-art Pearson and Spearman correlations with human subjective ratings. It consistently outperforms general MLLMs across speech, sound, music, and out-of-domain evaluation tasks without the need for task-specific retraining.

Jingxiang Chen, Minseok Kim, Seong-Gyun Leem et al.

Speech large language models (LLMs) observe paralinguistic cues such as prosody, emotion, and non-verbal sounds--crucial for intent understanding. However, leveraging these cues faces challenges: limited training data, annotation difficulty, and models exploiting lexical shortcuts over paralinguistic signals. We propose multi-task reinforcement learning (RL) with chain-of-thought prompting that elicits explicit affective reasoning. To address data scarcity, we introduce a paralinguistics-aware speech LLM (PALLM) that jointly optimizes sentiment classification from audio and paralinguistics-aware response generation via a two-stage pipeline. Experiments demonstrate that our approach improves paralinguistics understanding over both supervised baselines and strong proprietary models (Gemini-2.5-Pro, GPT-4o-audio) by 8-12% on Expresso, IEMOCAP, and RAVDESS. The results show that modeling paralinguistic reasoning with multi-task RL is crucial for building emotionally intelligent speech LLMs.

Chien-yu Huang, Wei-Chih Chen, Shu-wen Yang et al. · cmu, mit

Multimodal foundation models, such as Gemini and ChatGPT, have revolutionized human-machine interactions by seamlessly integrating various forms of data. Developing a universal spoken language model that comprehends a wide range of natural language instructions is critical for bridging communication gaps and facilitating more intuitive interactions. However, the absence of a comprehensive evaluation benchmark poses a significant challenge. We present Dynamic-SUPERB Phase-2, an open and evolving benchmark for the comprehensive evaluation of instruction-based universal speech models. Building upon the first generation, this second version incorporates 125 new tasks contributed collaboratively by the global research community, expanding the benchmark to a total of 180 tasks, making it the largest benchmark for speech and audio evaluation. While the first generation of Dynamic-SUPERB was limited to classification tasks, Dynamic-SUPERB Phase-2 broadens its evaluation capabilities by introducing a wide array of novel and diverse tasks, including regression and sequence generation, across speech, music, and environmental audio. Evaluation results show that no model performed well universally. SALMONN-13B excelled in English ASR and Qwen2-Audio-7B-Instruct showed high accuracy in emotion recognition, but current models still require further innovations to handle a broader range of tasks. We open-source all task data and the evaluation pipeline at https://github.com/dynamic-superb/dynamic-superb.

Andy T. Liu, Yi-Cheng Lin, Haibin Wu et al.

Despite their impressive success, training foundation models remains computationally costly. This paper investigates how to efficiently train speech foundation models with self-supervised learning (SSL) under a limited compute budget. We examine critical factors in SSL that impact the budget, including model architecture, model size, and data size. Our goal is to make analytical steps toward understanding the training dynamics of speech foundation models. We benchmark SSL objectives in an entirely comparable setting and find that other factors contribute more significantly to the success of SSL. Our results show that slimmer model architectures outperform common small architectures under the same compute and parameter budget. We demonstrate that the size of the pre-training data remains crucial, even with data augmentation during SSL training, as performance suffers when iterating over limited data. Finally, we identify a trade-off between model size and data size, highlighting an optimal model size for a given compute budget.

Haibin Wu, Po-chun Hsu, Ji Gao et al.

Automatic speaker verification (ASV), one of the most important technology for biometric identification, has been widely adopted in security-critical applications. However, ASV is seriously vulnerable to recently emerged adversarial attacks, yet effective countermeasures against them are limited. In this paper, we adopt neural vocoders to spot adversarial samples for ASV. We use the neural vocoder to re-synthesize audio and find that the difference between the ASV scores for the original and re-synthesized audio is a good indicator for discrimination between genuine and adversarial samples. This effort is, to the best of our knowledge, among the first to pursue such a technical direction for detecting time-domain adversarial samples for ASV, and hence there is a lack of established baselines for comparison. Consequently, we implement the Griffin-Lim algorithm as the detection baseline. The proposed approach achieves effective detection performance that outperforms the baselines in all the settings. We also show that the neural vocoder adopted in the detection framework is dataset-independent. Our codes will be made open-source for future works to do fair comparison.

Jian Hu, Siyang Jiang, Seth Austin Harding et al.

Many complex multi-agent systems such as robot swarms control and autonomous vehicle coordination can be modeled as Multi-Agent Reinforcement Learning (MARL) tasks. QMIX, a widely popular MARL algorithm, has been used as a baseline for the benchmark environments, e.g., Starcraft Multi-Agent Challenge (SMAC), Difficulty-Enhanced Predator-Prey (DEPP). Recent variants of QMIX target relaxing the monotonicity constraint of QMIX, allowing for performance improvement in SMAC. In this paper, we investigate the code-level optimizations of these variants and the monotonicity constraint. (1) We find that such improvements of the variants are significantly affected by various code-level optimizations. (2) The experiment results show that QMIX with normalized optimizations outperforms other works in SMAC; (3) beyond the common wisdom from these works, the monotonicity constraint can improve sample efficiency in SMAC and DEPP. We also discuss why monotonicity constraints work well in purely cooperative tasks with a theoretical analysis. We open-source the code at \url{https://github.com/hijkzzz/pymarl2}.

Shu-wen Yang, Heng-Jui Chang, Zili Huang et al. · meta-ai, mit

The foundation model paradigm leverages a shared foundation model to achieve state-of-the-art (SOTA) performance for various tasks, requiring minimal downstream-specific modeling and data annotation. This approach has proven crucial in the field of Natural Language Processing (NLP). However, the speech processing community lacks a similar setup to explore the paradigm systematically. In this work, we establish the Speech processing Universal PERformance Benchmark (SUPERB) to study the effectiveness of the paradigm for speech. We propose a unified multi-tasking framework to address speech processing tasks in SUPERB using a frozen foundation model followed by task-specialized, lightweight prediction heads. Combining our results with community submissions, we verify that the foundation model paradigm is promising for speech, and our multi-tasking framework is simple yet effective, as the best-performing foundation model shows competitive generalizability across most SUPERB tasks. For reproducibility and extensibility, we have developed a long-term maintained platform that enables deterministic benchmarking, allows for result sharing via an online leaderboard, and promotes collaboration through a community-driven benchmark database to support new development cycles. Finally, we conduct a series of analyses to offer an in-depth understanding of SUPERB and speech foundation models, including information flows across tasks inside the models, the correctness of the weighted-sum benchmarking protocol and the statistical significance and robustness of the benchmark.

Siddhant Arora, Kai-Wei Chang, Chung-Ming Chien et al. · mit, nvidia

The field of spoken language processing is undergoing a shift from training custom-built, task-specific models toward using and optimizing spoken language models (SLMs) which act as universal speech processing systems. This trend is similar to the progression toward universal language models that has taken place in the field of (text) natural language processing. SLMs include both "pure" language models of speech -- models of the distribution of tokenized speech sequences -- and models that combine speech encoders with text language models, often including both spoken and written input or output. Work in this area is very diverse, with a range of terminology and evaluation settings. This paper aims to contribute an improved understanding of SLMs via a unifying literature survey of recent work in the context of the evolution of the field. Our survey categorizes the work in this area by model architecture, training, and evaluation choices, and describes some key challenges and directions for future work.

Pooneh Mousavi, Gallil Maimon, Adel Moumen et al.

Discrete audio tokens are compact representations that aim to preserve perceptual quality, phonetic content, and speaker characteristics while enabling efficient storage and inference, as well as competitive performance across diverse downstream tasks. They provide a practical alternative to continuous features, enabling the integration of speech and audio into modern large language models (LLMs). As interest in token-based audio processing grows, various tokenization methods have emerged, and several surveys have reviewed the latest progress in the field. However, existing studies often focus on specific domains or tasks and lack a unified comparison across various benchmarks. This paper presents a systematic review and benchmark of discrete audio tokenizers, covering three domains: speech, music, and general audio. We propose a taxonomy of tokenization approaches based on encoder-decoder, quantization techniques, training paradigm, streamability, and application domains. We evaluate tokenizers on multiple benchmarks for reconstruction, downstream performance, and acoustic language modeling, and analyze trade-offs through controlled ablation studies. Our findings highlight key limitations, practical considerations, and open challenges, providing insight and guidance for future research in this rapidly evolving area. For more information, including our main results and tokenizer database, please refer to our website: https://poonehmousavi.github.io/dates-website/.

Yifan Yang, Shujie Liu, Jinyu Li et al.

Recent zero-shot text-to-speech (TTS) systems face a common dilemma: autoregressive (AR) models suffer from slow generation and lack duration controllability, while non-autoregressive (NAR) models lack temporal modeling and typically require complex designs. In this paper, we introduce a novel pseudo-autoregressive (PAR) codec language modeling approach that unifies AR and NAR modeling. Combining explicit temporal modeling from AR with parallel generation from NAR, PAR generates dynamic-length spans at fixed time steps. Building on PAR, we propose PALLE, a two-stage TTS system that leverages PAR for initial generation followed by NAR refinement. In the first stage, PAR progressively generates speech tokens along the time dimension, with each step predicting all positions in parallel but only retaining the left-most span. In the second stage, low-confidence tokens are iteratively refined in parallel, leveraging the global contextual information. Experiments demonstrate that PALLE, trained on LibriTTS, outperforms state-of-the-art systems trained on large-scale data, including F5-TTS, E2-TTS, and MaskGCT, on the LibriSpeech test-clean set in terms of speech quality, speaker similarity, and intelligibility, while achieving up to ten times faster inference speed. Audio samples are available at https://microsoft.com/research/project/vall-e-x/palle.

Haibin Wu, Heng-Cheng Kuo, Yu Tsao et al.

Automatic speaker verification (ASV) is highly susceptible to adversarial attacks. Purification modules are usually adopted as a pre-processing to mitigate adversarial noise. However, they are commonly implemented across diverse experimental settings, rendering direct comparisons challenging. This paper comprehensively compares mainstream purification techniques in a unified framework. We find these methods often face a trade-off between user experience and security, as they struggle to simultaneously maintain genuine sample performance and reduce adversarial perturbations. To address this challenge, some efforts have extended purification modules to encompass detection capabilities, aiming to alleviate the trade-off. However, advanced purification modules will always come into the stage to surpass previous detection method. As a result, we further propose an easy-to-follow ensemble approach that integrates advanced purification modules for detection, achieving state-of-the-art (SOTA) performance in countering adversarial noise. Our ensemble method has great potential due to its compatibility with future advanced purification techniques.

Leying Zhang, Yao Qian, Xiaofei Wang et al.

Generating natural-sounding, multi-speaker dialogue is crucial for applications such as podcast creation, virtual agents, and multimedia content generation. However, existing systems struggle to maintain speaker consistency, model overlapping speech, and synthesize coherent conversations efficiently. In this paper, we introduce CoVoMix2, a fully non-autoregressive framework for zero-shot multi-talker dialogue generation. CoVoMix2 directly predicts mel-spectrograms from multi-stream transcriptions using a flow-matching-based generative model, eliminating the reliance on intermediate token representations. To better capture realistic conversational dynamics, we propose transcription-level speaker disentanglement, sentence-level alignment, and prompt-level random masking strategies. Our approach achieves state-of-the-art performance, outperforming strong baselines like MoonCast and Sesame in speech quality, speaker consistency, and inference speed. Notably, CoVoMix2 operates without requiring transcriptions for the prompt and supports controllable dialogue generation, including overlapping speech and precise timing control, demonstrating strong generalizability to real-world speech generation scenarios.

Xuanjun Chen, Chia-Yu Hu, Sung-Feng Huang et al.

Rapid advances in singing voice synthesis have increased unauthorized imitation risks, creating an urgent need for better Singing Voice Deepfake (SingFake) Detection, also known as SVDD. Unlike speech, singing contains complex pitch, wide dynamic range, and timbral variations. Conventional 16 kHz-sampled detectors prove inadequate, as they discard vital high-frequency information. This study presents the first systematic analysis of high-resolution (44.1 kHz sampling rate) audio for SVDD. We propose a joint fullband-subband modeling framework: the fullband captures global context, while subband-specific experts isolate fine-grained synthesis artifacts unevenly distributed across the spectrum. Experiments on the WildSVDD dataset demonstrate that high-frequency subbands provide essential complementary cues. Our framework significantly outperforms 16 kHz-sampled models, proving that high-resolution audio and strategic subband integration are critical for robust in-the-wild detection.

Xuanjun Chen, Shih-Peng Cheng, Jiawei Du et al.

Audio-visual temporal deepfake localization under the content-driven partial manipulation remains a highly challenging task. In this scenario, the deepfake regions are usually only spanning a few frames, with the majority of the rest remaining identical to the original. To tackle this, we propose a Hierarchical Boundary Modeling Network (HBMNet), which includes three modules: an Audio-Visual Feature Encoder that extracts discriminative frame-level representations, a Coarse Proposal Generator that predicts candidate boundary regions, and a Fine-grained Probabilities Generator that refines these proposals using bidirectional boundary-content probabilities. From the modality perspective, we enhance audio-visual learning through dedicated encoding and fusion, reinforced by frame-level supervision to boost discriminability. From the temporal perspective, HBMNet integrates multi-scale cues and bidirectional boundary-content relationships. Experiments show that encoding and fusion primarily improve precision, while frame-level supervision boosts recall. Each module (audio-visual fusion, temporal scales, bi-directionality) contributes complementary benefits, collectively enhancing localization performance. HBMNet outperforms BA-TFD and UMMAFormer and shows improved potential scalability with more training data.

Xuanjun Chen, I-Ming Lin, Lin Zhang et al.

Recent attempts at source tracing for codec-based deepfake speech (CodecFake), generated by neural audio codec-based speech generation (CoSG) models, have exhibited suboptimal performance. However, how to train source tracing models using simulated CoSG data while maintaining strong performance on real CoSG-generated audio remains an open challenge. In this paper, we show that models trained solely on codec-resynthesized data tend to overfit to non-speech regions and struggle to generalize to unseen content. To mitigate these challenges, we introduce the Semantic-Acoustic Source Tracing Network (SASTNet), which jointly leverages Whisper for semantic feature encoding and Wav2vec2 with AudioMAE for acoustic feature encoding. Our proposed SASTNet achieves state-of-the-art performance on the CoSG test set of the CodecFake+ dataset, demonstrating its effectiveness for reliable source tracing.

Haibin Wu, Yuxuan Hu, Ruchao Fan et al.

Speech language models (Speech LMs) enable end-to-end speech-text modelling within a single model, offering a promising direction for spoken dialogue systems. The choice of speech-text jointly decoding paradigm plays a critical role in performance, efficiency, and alignment quality. In this work, we systematically compare representative joint speech-text decoding strategies-including the interleaved, and parallel generation paradigms-under a controlled experimental setup using the same base language model, speech tokenizer and training data. Our results show that the interleaved approach achieves the best alignment. However it suffers from slow inference due to long token sequence length. To address this, we propose a novel early-stop interleaved (ESI) pattern that not only significantly accelerates decoding but also yields slightly better performance. Additionally, we curate high-quality question answering (QA) datasets to further improve speech QA performance.

Lingwei Meng, Jiawen Kang, Mingyu Cui et al.

Multi-talker overlapped speech poses a significant challenge for speech recognition and diarization. Recent research indicated that these two tasks are inter-dependent and complementary, motivating us to explore a unified modeling method to address them in the context of overlapped speech. A recent study proposed a cost-effective method to convert a single-talker automatic speech recognition (ASR) system into a multi-talker one, by inserting a Sidecar separator into the frozen well-trained ASR model. Extending on this, we incorporate a diarization branch into the Sidecar, allowing for unified modeling of both ASR and diarization with a negligible overhead of only 768 parameters. The proposed method yields better ASR results compared to the baseline on LibriMix and LibriSpeechMix datasets. Moreover, without sophisticated customization on the diarization task, our method achieves acceptable diarization results on the two-speaker subset of CALLHOME with only a few adaptation steps.

Haibin Wu, Jiawen Kang, Lingwei Meng et al.

Automatic speaker verification (ASV) plays a critical role in security-sensitive environments. Regrettably, the reliability of ASV has been undermined by the emergence of spoofing attacks, such as replay and synthetic speech, as well as adversarial attacks and the relatively new partially fake speech. While there are several review papers that cover replay and synthetic speech, and adversarial attacks, there is a notable gap in a comprehensive review that addresses defense against adversarial attacks and the recently emerged partially fake speech. Thus, the aim of this paper is to provide a thorough and systematic overview of the defense methods used against these types of attacks.

Haibin Wu, Heng-Cheng Kuo, Naijun Zheng et al.

The past few years have witnessed the significant advances of speech synthesis and voice conversion technologies. However, such technologies can undermine the robustness of broadly implemented biometric identification models and can be harnessed by in-the-wild attackers for illegal uses. The ASVspoof challenge mainly focuses on synthesized audios by advanced speech synthesis and voice conversion models, and replay attacks. Recently, the first Audio Deep Synthesis Detection challenge (ADD 2022) extends the attack scenarios into more aspects. Also ADD 2022 is the first challenge to propose the partially fake audio detection task. Such brand new attacks are dangerous and how to tackle such attacks remains an open question. Thus, we propose a novel framework by introducing the question-answering (fake span discovery) strategy with the self-attention mechanism to detect partially fake audios. The proposed fake span detection module tasks the anti-spoofing model to predict the start and end positions of the fake clip within the partially fake audio, address the model's attention into discovering the fake spans rather than other shortcuts with less generalization, and finally equips the model with the discrimination capacity between real and partially fake audios. Our submission ranked second in the partially fake audio detection track of ADD 2022.

Naijun Zheng, Na Li, Xixin Wu et al.

This paper describes our speaker diarization system submitted to the Multi-channel Multi-party Meeting Transcription (M2MeT) challenge, where Mandarin meeting data were recorded in multi-channel format for diarization and automatic speech recognition (ASR) tasks. In these meeting scenarios, the uncertainty of the speaker number and the high ratio of overlapped speech present great challenges for diarization. Based on the assumption that there is valuable complementary information between acoustic features, spatial-related and speaker-related features, we propose a multi-level feature fusion mechanism based target-speaker voice activity detection (FFM-TS-VAD) system to improve the performance of the conventional TS-VAD system. Furthermore, we propose a data augmentation method during training to improve the system robustness when the angular difference between two speakers is relatively small. We provide comparisons for different sub-systems we used in M2MeT challenge. Our submission is a fusion of several sub-systems and ranks second in the diarization task.

Haibin Wu, Bo Zheng, Xu Li et al.

A leaderboard named Speech processing Universal PERformance Benchmark (SUPERB), which aims at benchmarking the performance of a shared self-supervised learning (SSL) speech model across various downstream speech tasks with minimal modification of architectures and small amount of data, has fueled the research for speech representation learning. The SUPERB demonstrates speech SSL upstream models improve the performance of various downstream tasks through just minimal adaptation. As the paradigm of the self-supervised learning upstream model followed by downstream tasks arouses more attention in the speech community, characterizing the adversarial robustness of such paradigm is of high priority. In this paper, we make the first attempt to investigate the adversarial vulnerability of such paradigm under the attacks from both zero-knowledge adversaries and limited-knowledge adversaries. The experimental results illustrate that the paradigm proposed by SUPERB is seriously vulnerable to limited-knowledge adversaries, and the attacks generated by zero-knowledge adversaries are with transferability. The XAB test verifies the imperceptibility of crafted adversarial attacks.

Haibin Wu, Yang Zhang, Zhiyong Wu et al.

Automatic speaker verification (ASV) is a well developed technology for biometric identification, and has been ubiquitous implemented in security-critic applications, such as banking and access control. However, previous works have shown that ASV is under the radar of adversarial attacks, which are very similar to their original counterparts from human's perception, yet will manipulate the ASV render wrong prediction. Due to the very late emergence of adversarial attacks for ASV, effective countermeasures against them are limited. Given that the security of ASV is of high priority, in this work, we propose the idea of "voting for the right answer" to prevent risky decisions of ASV in blind spot areas, by employing random sampling and voting. Experimental results show that our proposed method improves the robustness against both the limited-knowledge attackers by pulling the adversarial samples out of the blind spots, and the perfect-knowledge attackers by introducing randomness and increasing the attackers' budgets.

Haibin Wu, Xu Li, Andy T. Liu et al.

Previous works have shown that automatic speaker verification (ASV) is seriously vulnerable to malicious spoofing attacks, such as replay, synthetic speech, and recently emerged adversarial attacks. Great efforts have been dedicated to defending ASV against replay and synthetic speech; however, only a few approaches have been explored to deal with adversarial attacks. All the existing approaches to tackle adversarial attacks for ASV require the knowledge for adversarial samples generation, but it is impractical for defenders to know the exact attack algorithms that are applied by the in-the-wild attackers. This work is among the first to perform adversarial defense for ASV without knowing the specific attack algorithms. Inspired by self-supervised learning models (SSLMs) that possess the merits of alleviating the superficial noise in the inputs and reconstructing clean samples from the interrupted ones, this work regards adversarial perturbations as one kind of noise and conducts adversarial defense for ASV by SSLMs. Specifically, we propose to perform adversarial defense from two perspectives: 1) adversarial perturbation purification and 2) adversarial perturbation detection. Experimental results show that our detection module effectively shields the ASV by detecting adversarial samples with an accuracy of around 80%. Moreover, since there is no common metric for evaluating the adversarial defense performance for ASV, this work also formalizes evaluation metrics for adversarial defense considering both purification and detection based approaches into account. We sincerely encourage future works to benchmark their approaches based on the proposed evaluation framework.

Yuqiong Qi, Yang Hu, Haibin Wu et al.

The state-of-the-art driving automation system demands extreme computational resources to meet rigorous accuracy and latency requirements. Though emerging driving automation computing platforms are based on ASIC to provide better performance and power guarantee, building such an accelerator-based computing platform for driving automation still present challenges. First, the workloads mix and performance requirements exposed to driving automation system present significant variability. Second, with more cameras/sensors integrated in a future fully autonomous driving vehicle, a heterogeneous multi-accelerator architecture substrate is needed that requires a design space exploration for a new form of parallelism. In this work, we aim to extensively explore the above system design challenges and these challenges motivate us to propose a comprehensive framework that synergistically handles the heterogeneous hardware accelerator design principles, system design criteria, and task scheduling mechanism. Specifically, we propose a novel heterogeneous multi-core AI accelerator (HMAI) to provide the hardware substrate for the driving automation tasks with variability. We also define system design criteria to better utilize hardware resources and achieve increased throughput while satisfying the performance and energy restrictions. Finally, we propose a deep reinforcement learning (RL)-based task scheduling mechanism FlexAI, to resolve task mapping issue. Experimental results show that with FlexAI scheduling, basically 100% tasks in each driving route can be processed by HMAI within their required period to ensure safety, and FlexAI can also maximally reduce the breaking distance up to 96% as compared to typical heuristics and guided random-search-based algorithms.

Haibin Wu, Xu Li, Andy T. Liu et al.

Automatic speaker verification (ASV) is one of the core technologies in biometric identification. With the ubiquitous usage of ASV systems in safety-critical applications, more and more malicious attackers attempt to launch adversarial attacks at ASV systems. In the midst of the arms race between attack and defense in ASV, how to effectively improve the robustness of ASV against adversarial attacks remains an open question. We note that the self-supervised learning models possess the ability to mitigate superficial perturbations in the input after pretraining. Hence, with the goal of effective defense in ASV against adversarial attacks, we propose a standard and attack-agnostic method based on cascaded self-supervised learning models to purify the adversarial perturbations. Experimental results demonstrate that the proposed method achieves effective defense performance and can successfully counter adversarial attacks in scenarios where attackers may either be aware or unaware of the self-supervised learning models.

Jian Hu, Seth Austin Harding, Haibin Wu et al.

In Cooperative Multi-Agent Reinforcement Learning (MARL) and under the setting of Centralized Training with Decentralized Execution (CTDE), agents observe and interact with their environment locally and independently. With local observation and random sampling, the randomness in rewards and observations leads to randomness in long-term returns. Existing methods such as Value Decomposition Network (VDN) and QMIX estimate the value of long-term returns as a scalar that does not contain the information of randomness. Our proposed model QR-MIX introduces quantile regression, modeling joint state-action values as a distribution, combining QMIX with Implicit Quantile Network (IQN). However, the monotonicity in QMIX limits the expression of joint state-action value distribution and may lead to incorrect estimation results in non-monotonic cases. Therefore, we proposed a flexible loss function to approximate the monotonicity found in QMIX. Our model is not only more tolerant of the randomness of returns, but also more tolerant of the randomness of monotonic constraints. The experimental results demonstrate that QR-MIX outperforms the previous state-of-the-art method QMIX in the StarCraft Multi-Agent Challenge (SMAC) environment.

Haibin Wu, Andy T. Liu, Hung-yi Lee

High-performance anti-spoofing models for automatic speaker verification (ASV), have been widely used to protect ASV by identifying and filtering spoofing audio that is deliberately generated by text-to-speech, voice conversion, audio replay, etc. However, it has been shown that high-performance anti-spoofing models are vulnerable to adversarial attacks. Adversarial attacks, that are indistinguishable from original data but result in the incorrect predictions, are dangerous for anti-spoofing models and not in dispute we should detect them at any cost. To explore this issue, we proposed to employ Mockingjay, a self-supervised learning based model, to protect anti-spoofing models against adversarial attacks in the black-box scenario. Self-supervised learning models are effective in improving downstream task performance like phone classification or ASR. However, their effect in defense for adversarial attacks has not been explored yet. In this work, we explore the robustness of self-supervised learned high-level representations by using them in the defense against adversarial attacks. A layerwise noise to signal ratio (LNSR) is proposed to quantize and measure the effectiveness of deep models in countering adversarial noise. Experimental results on the ASVspoof 2019 dataset demonstrate that high-level representations extracted by Mockingjay can prevent the transferability of adversarial examples, and successfully counter black-box attacks.

Haibin Wu, Songxiang Liu, Helen Meng et al.

Various forefront countermeasure methods for automatic speaker verification (ASV) with considerable performance in anti-spoofing are proposed in the ASVspoof 2019 challenge. However, previous work has shown that countermeasure models are vulnerable to adversarial examples indistinguishable from natural data. A good countermeasure model should not only be robust against spoofing audio, including synthetic, converted, and replayed audios; but counteract deliberately generated examples by malicious adversaries. In this work, we introduce a passive defense method, spatial smoothing, and a proactive defense method, adversarial training, to mitigate the vulnerability of ASV spoofing countermeasure models against adversarial examples. This paper is among the first to use defense methods to improve the robustness of ASV spoofing countermeasure models under adversarial attacks. The experimental results show that these two defense methods positively help spoofing countermeasure models counter adversarial examples.

Songxiang Liu, Haibin Wu, Hung-yi Lee et al.

High-performance spoofing countermeasure systems for automatic speaker verification (ASV) have been proposed in the ASVspoof 2019 challenge. However, the robustness of such systems under adversarial attacks has not been studied yet. In this paper, we investigate the vulnerability of spoofing countermeasures for ASV under both white-box and black-box adversarial attacks with the fast gradient sign method (FGSM) and the projected gradient descent (PGD) method. We implement high-performing countermeasure models in the ASVspoof 2019 challenge and conduct adversarial attacks on them. We compare performance of black-box attacks across spoofing countermeasure models with different network architectures and different amount of model parameters. The experimental results show that all implemented countermeasure models are vulnerable to FGSM and PGD attacks under the scenario of white-box attack. The more dangerous black-box attacks also prove to be effective by the experimental results.