Jiatong Shi

Lester Phillip Violeta, Xueyao Zhang, Jiatong Shi et al.

We present a thorough analysis of the findings of the latest iteration of the Singing Voice Conversion Challenge, a scientific event aiming to compare and understand different voice conversion systems in a controlled environment. Compared to previous iterations which solely focused on converting the singer identity, this year we also focused on converting the singing style of the singer. To create a controlled environment and thorough evaluations, we developed a new challenge database, introduced two tasks, open-sourced baselines, and conducted large-scale crowd-sourced listening tests and objective evaluations. The challenge was run for two months and in total we evaluated 33 different systems. The results of the large-scale crowd-sourced listening test showed that top systems had comparable singer identity scores to ground truth samples. However, modeling the singing style and consequently achieving high naturalness still remains a challenge in this task, primarily due to the difficulty in modeling dynamic information in breathy, glissando, and vibrato singing styles. Further analyses of the challenge also discuss the limitations of both the traditional similarity test and the dynamic preference test in evaluating singing style similarity. Moreover, calculating Spearman's rank correlation coefficient shows that dependent objective metrics such as chroma-alignment and non-match metrics such as speaker embeddings are the most correlated to subjective scores, but are still not at a level where it could be considered as a true replacement for subjective scores.

Yifan Peng, Jinchuan Tian, Brian Yan et al. · cmu, meta-ai

Pre-training speech models on large volumes of data has achieved remarkable success. OpenAI Whisper is a multilingual multitask model trained on 680k hours of supervised speech data. It generalizes well to various speech recognition and translation benchmarks even in a zero-shot setup. However, the full pipeline for developing such models (from data collection to training) is not publicly accessible, which makes it difficult for researchers to further improve its performance and address training-related issues such as efficiency, robustness, fairness, and bias. This work presents an Open Whisper-style Speech Model (OWSM), which reproduces Whisper-style training using an open-source toolkit and publicly available data. OWSM even supports more translation directions and can be more efficient to train. We will publicly release all scripts used for data preparation, training, inference, and scoring as well as pre-trained models and training logs to promote open science.

Brian Yan, Jiatong Shi, Yun Tang et al. · cmu, nvidia

ESPnet-ST-v2 is a revamp of the open-source ESPnet-ST toolkit necessitated by the broadening interests of the spoken language translation community. ESPnet-ST-v2 supports 1) offline speech-to-text translation (ST), 2) simultaneous speech-to-text translation (SST), and 3) offline speech-to-speech translation (S2ST) -- each task is supported with a wide variety of approaches, differentiating ESPnet-ST-v2 from other open source spoken language translation toolkits. This toolkit offers state-of-the-art architectures such as transducers, hybrid CTC/attention, multi-decoders with searchable intermediates, time-synchronous blockwise CTC/attention, Translatotron models, and direct discrete unit models. In this paper, we describe the overall design, example models for each task, and performance benchmarking behind ESPnet-ST-v2, which is publicly available at https://github.com/espnet/espnet.

William Chen, Brian Yan, Jiatong Shi et al. · nvidia

Multilingual Automatic Speech Recognition (ASR) models have extended the usability of speech technologies to a wide variety of languages. With how many languages these models have to handle, however, a key to understanding their imbalanced performance across different languages is to examine if the model actually knows which language it should transcribe. In this paper, we introduce our work on improving performance on FLEURS, a 102-language open ASR benchmark, by conditioning the entire model on language identity (LID). We investigate techniques inspired from recent Connectionist Temporal Classification (CTC) studies to help the model handle the large number of languages, conditioning on the LID predictions of auxiliary tasks. Our experimental results demonstrate the effectiveness of our technique over standard CTC/Attention-based hybrid models. Furthermore, our state-of-the-art systems using self-supervised models with the Conformer architecture improve over the results of prior work on FLEURS by a relative 28.4% CER. Trained models and reproducible recipes are available at https://github.com/espnet/espnet/tree/master/egs2/fleurs/asr1 .

Rongjie Huang, Mingze Li, Dongchao Yang et al.

Large language models (LLMs) have exhibited remarkable capabilities across a variety of domains and tasks, challenging our understanding of learning and cognition. Despite the recent success, current LLMs are not capable of processing complex audio information or conducting spoken conversations (like Siri or Alexa). In this work, we propose a multi-modal AI system named AudioGPT, which complements LLMs (i.e., ChatGPT) with 1) foundation models to process complex audio information and solve numerous understanding and generation tasks; and 2) the input/output interface (ASR, TTS) to support spoken dialogue. With an increasing demand to evaluate multi-modal LLMs of human intention understanding and cooperation with foundation models, we outline the principles and processes and test AudioGPT in terms of consistency, capability, and robustness. Experimental results demonstrate the capabilities of AudioGPT in solving AI tasks with speech, music, sound, and talking head understanding and generation in multi-round dialogues, which empower humans to create rich and diverse audio content with unprecedented ease. Our system is publicly available at \url{https://github.com/AIGC-Audio/AudioGPT}.

William Chen, Jiatong Shi, Brian Yan et al. · nvidia

Multilingual self-supervised learning (SSL) has often lagged behind state-of-the-art (SOTA) methods due to the expenses and complexity required to handle many languages. This further harms the reproducibility of SSL, which is already limited to few research groups due to its resource usage. We show that more powerful techniques can actually lead to more efficient pre-training, opening SSL to more research groups. We propose WavLabLM, which extends WavLM's joint prediction and denoising to 40k hours of data across 136 languages. To build WavLabLM, we devise a novel multi-stage pre-training method, designed to address the language imbalance of multilingual data. WavLabLM achieves comparable performance to XLS-R on ML-SUPERB with less than 10% of the training data, making SSL realizable with academic compute. We show that further efficiency can be achieved with a vanilla HuBERT Base model, which can maintain 94% of XLS-R's performance with only 3% of the data, 4 GPUs, and limited trials. We open-source all code and models in ESPnet.

Hsiang-Sheng Tsai, Heng-Jui Chang, Wen-Chin Huang et al. · meta-ai, mit

Transfer learning has proven to be crucial in advancing the state of speech and natural language processing research in recent years. In speech, a model pre-trained by self-supervised learning transfers remarkably well on multiple tasks. However, the lack of a consistent evaluation methodology is limiting towards a holistic understanding of the efficacy of such models. SUPERB was a step towards introducing a common benchmark to evaluate pre-trained models across various speech tasks. In this paper, we introduce SUPERB-SG, a new benchmark focused on evaluating the semantic and generative capabilities of pre-trained models by increasing task diversity and difficulty over SUPERB. We use a lightweight methodology to test the robustness of representations learned by pre-trained models under shifts in data domain and quality across different types of tasks. It entails freezing pre-trained model parameters, only using simple task-specific trainable heads. The goal is to be inclusive of all researchers, and encourage efficient use of computational resources. We also show that the task diversity of SUPERB-SG coupled with limited task supervision is an effective recipe for evaluating the generalizability of model representation.

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.

Xuankai Chang, Brian Yan, Kwanghee Choi et al. · cmu, meta-ai

Speech signals, typically sampled at rates in the tens of thousands per second, contain redundancies, evoking inefficiencies in sequence modeling. High-dimensional speech features such as spectrograms are often used as the input for the subsequent model. However, they can still be redundant. Recent investigations proposed the use of discrete speech units derived from self-supervised learning representations, which significantly compresses the size of speech data. Applying various methods, such as de-duplication and subword modeling, can further compress the speech sequence length. Hence, training time is significantly reduced while retaining notable performance. In this study, we undertake a comprehensive and systematic exploration into the application of discrete units within end-to-end speech processing models. Experiments on 12 automatic speech recognition, 3 speech translation, and 1 spoken language understanding corpora demonstrate that discrete units achieve reasonably good results in almost all the settings. We intend to release our configurations and trained models to foster future research efforts.

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.

Kalvin Chang, Yi-Hui Chou, Jiatong Shi et al. · cmu

Underperformance of ASR systems for speakers of African American Vernacular English (AAVE) and other marginalized language varieties is a well-documented phenomenon, and one that reinforces the stigmatization of these varieties. We investigate whether or not the recent wave of Self-Supervised Learning (SSL) speech models can close the gap in ASR performance between AAVE and Mainstream American English (MAE). We evaluate four SSL models (wav2vec 2.0, HuBERT, WavLM, and XLS-R) on zero-shot Automatic Speech Recognition (ASR) for these two varieties and find that these models perpetuate the bias in performance against AAVE. Additionally, the models have higher word error rates on utterances with more phonological and morphosyntactic features of AAVE. Despite the success of SSL speech models in improving ASR for low resource varieties, SSL pre-training alone may not bridge the gap between AAVE and MAE. Our code is publicly available at https://github.com/cmu-llab/s3m-aave.

Jiatong Shi, Yun Tang, Ann Lee et al. · meta-ai

It has been known that direct speech-to-speech translation (S2ST) models usually suffer from the data scarcity issue because of the limited existing parallel materials for both source and target speech. Therefore to train a direct S2ST system, previous works usually utilize text-to-speech (TTS) systems to generate samples in the target language by augmenting the data from speech-to-text translation (S2TT). However, there is a limited investigation into how the synthesized target speech would affect the S2ST models. In this work, we analyze the effect of changing synthesized target speech for direct S2ST models. We find that simply combining the target speech from different TTS systems can potentially improve the S2ST performances. Following that, we also propose a multi-task framework that jointly optimizes the S2ST system with multiple targets from different TTS systems. Extensive experiments demonstrate that our proposed framework achieves consistent improvements (2.8 BLEU) over the baselines on the Fisher Spanish-English dataset.

Dongji Gao, Jiatong Shi, Shun-Po Chuang et al.

This paper describes the ESPnet Unsupervised ASR Open-source Toolkit (EURO), an end-to-end open-source toolkit for unsupervised automatic speech recognition (UASR). EURO adopts the state-of-the-art UASR learning method introduced by the Wav2vec-U, originally implemented at FAIRSEQ, which leverages self-supervised speech representations and adversarial training. In addition to wav2vec2, EURO extends the functionality and promotes reproducibility for UASR tasks by integrating S3PRL and k2, resulting in flexible frontends from 27 self-supervised models and various graph-based decoding strategies. EURO is implemented in ESPnet and follows its unified pipeline to provide UASR recipes with a complete setup. This improves the pipeline's efficiency and allows EURO to be easily applied to existing datasets in ESPnet. Extensive experiments on three mainstream self-supervised models demonstrate the toolkit's effectiveness and achieve state-of-the-art UASR performance on TIMIT and LibriSpeech datasets. EURO will be publicly available at https://github.com/espnet/espnet, aiming to promote this exciting and emerging research area based on UASR through open-source activity.

Masao Someki, Kwanghee Choi, Siddhant Arora et al. · cmu, nvidia

We introduce ESPnet-EZ, an extension of the open-source speech processing toolkit ESPnet, aimed at quick and easy development of speech models. ESPnet-EZ focuses on two major aspects: (i) easy fine-tuning and inference of existing ESPnet models on various tasks and (ii) easy integration with popular deep neural network frameworks such as PyTorch-Lightning, Hugging Face transformers and datasets, and Lhotse. By replacing ESPnet design choices inherited from Kaldi with a Python-only, Bash-free interface, we dramatically reduce the effort required to build, debug, and use a new model. For example, to fine-tune a speech foundation model, ESPnet-EZ, compared to ESPnet, reduces the number of newly written code by 2.7x and the amount of dependent code by 6.7x while dramatically reducing the Bash script dependencies. The codebase of ESPnet-EZ is publicly available.

Jiatong Shi, William Chen, Dan Berrebbi et al. · meta-ai, mit

The 2023 Multilingual Speech Universal Performance Benchmark (ML-SUPERB) Challenge expands upon the acclaimed SUPERB framework, emphasizing self-supervised models in multilingual speech recognition and language identification. The challenge comprises a research track focused on applying ML-SUPERB to specific multilingual subjects, a Challenge Track for model submissions, and a New Language Track where language resource researchers can contribute and evaluate their low-resource language data in the context of the latest progress in multilingual speech recognition. The challenge garnered 12 model submissions and 54 language corpora, resulting in a comprehensive benchmark encompassing 154 languages. The findings indicate that merely scaling models is not the definitive solution for multilingual speech tasks, and a variety of speech/voice types present significant challenges in multilingual speech processing.

Jiatong Shi, George Saon, David Haws et al. · ibm-research

Beam search, which is the dominant ASR decoding algorithm for end-to-end models, generates tree-structured hypotheses. However, recent studies have shown that decoding with hypothesis merging can achieve a more efficient search with comparable or better performance. But, the full context in recurrent networks is not compatible with hypothesis merging. We propose to use vector-quantized long short-term memory units (VQ-LSTM) in the prediction network of RNN transducers. By training the discrete representation jointly with the ASR network, hypotheses can be actively merged for lattice generation. Our experiments on the Switchboard corpus show that the proposed VQ RNN transducers improve ASR performance over transducers with regular prediction networks while also producing denser lattices with a very low oracle word error rate (WER) for the same beam size. Additional language model rescoring experiments also demonstrate the effectiveness of the proposed lattice generation scheme.

Dan Berrebbi, Jiatong Shi, Brian Yan et al.

Self-Supervised Learning (SSL) models have been successfully applied in various deep learning-based speech tasks, particularly those with a limited amount of data. However, the quality of SSL representations depends highly on the relatedness between the SSL training domain(s) and the target data domain. On the contrary, spectral feature (SF) extractors such as log Mel-filterbanks are hand-crafted non-learnable components, and could be more robust to domain shifts. The present work examines the assumption that combining non-learnable SF extractors to SSL models is an effective approach to low resource speech tasks. We propose a learnable and interpretable framework to combine SF and SSL representations. The proposed framework outperforms significantly both baseline and SSL models on Automatic Speech Recognition (ASR) and Speech Translation (ST) tasks on three low resource datasets. We additionally design a mixture of experts based combination model. This last model reveals that the relative contribution of SSL models over conventional SF extractors is very small in case of domain mismatch between SSL training set and the target language data.

Wen-Chin Huang, Lester Phillip Violeta, Songxiang Liu et al.

We present the latest iteration of the voice conversion challenge (VCC) series, a bi-annual scientific event aiming to compare and understand different voice conversion (VC) systems based on a common dataset. This year we shifted our focus to singing voice conversion (SVC), thus named the challenge the Singing Voice Conversion Challenge (SVCC). A new database was constructed for two tasks, namely in-domain and cross-domain SVC. The challenge was run for two months, and in total we received 26 submissions, including 2 baselines. Through a large-scale crowd-sourced listening test, we observed that for both tasks, although human-level naturalness was achieved by the top system, no team was able to obtain a similarity score as high as the target speakers. Also, as expected, cross-domain SVC is harder than in-domain SVC, especially in the similarity aspect. We also investigated whether existing objective measurements were able to predict perceptual performance, and found that only few of them could reach a significant correlation.

Yen Meng, Hsuan-Jui Chen, Jiatong Shi et al.

Compressing self-supervised models has become increasingly necessary, as self-supervised models become larger. While previous approaches have primarily focused on compressing the model size, shortening sequences is also effective in reducing the computational cost. In this work, we study fixed-length and variable-length subsampling along the time axis in self-supervised learning. We explore how individual downstream tasks are sensitive to input frame rates. Subsampling while training self-supervised models not only improves the overall performance on downstream tasks under certain frame rates, but also brings significant speed-up in inference. Variable-length subsampling performs particularly well under low frame rates. In addition, if we have access to phonetic boundaries, we find no degradation in performance for an average frame rate as low as 10 Hz.

Keqi Deng, Shinji Watanabe, Jiatong Shi et al.

Although Transformers have gained success in several speech processing tasks like spoken language understanding (SLU) and speech translation (ST), achieving online processing while keeping competitive performance is still essential for real-world interaction. In this paper, we take the first step on streaming SLU and simultaneous ST using a blockwise streaming Transformer, which is based on contextual block processing and blockwise synchronous beam search. Furthermore, we design an automatic speech recognition (ASR)-based intermediate loss regularization for the streaming SLU task to improve the classification performance further. As for the simultaneous ST task, we propose a cross-lingual encoding method, which employs a CTC branch optimized with target language translations. In addition, the CTC translation output is also used to refine the search space with CTC prefix score, achieving joint CTC/attention simultaneous translation for the first time. Experiments for SLU are conducted on FSC and SLURP corpora, while the ST task is evaluated on Fisher-CallHome Spanish and MuST-C En-De corpora. Experimental results show that the blockwise streaming Transformer achieves competitive results compared to offline models, especially with our proposed methods that further yield a 2.4% accuracy gain on the SLU task and a 4.3 BLEU gain on the ST task over streaming baselines.

Jiatong Shi, Chan-Jan Hsu, Holam Chung et al.

Spoken language understanding (SLU) is a task aiming to extract high-level semantics from spoken utterances. Previous works have investigated the use of speech self-supervised models and textual pre-trained models, which have shown reasonable improvements to various SLU tasks. However, because of the mismatched modalities between speech signals and text tokens, previous methods usually need complex designs of the frameworks. This work proposes a simple yet efficient unsupervised paradigm that connects speech and textual pre-trained models, resulting in an unsupervised speech-to-semantic pre-trained model for various tasks in SLU. To be specific, we propose to use unsupervised automatic speech recognition (ASR) as a connector that bridges different modalities used in speech and textual pre-trained models. Our experiments show that unsupervised ASR itself can improve the representations from speech self-supervised models. More importantly, it is shown as an efficient connector between speech and textual pre-trained models, improving the performances of five different SLU tasks. Notably, on spoken question answering, we reach the state-of-the-art result over the challenging NMSQA benchmark.

Yui Sudo, Muhammad Shakeel, Brian Yan et al.

The network architecture of end-to-end (E2E) automatic speech recognition (ASR) can be classified into several models, including connectionist temporal classification (CTC), recurrent neural network transducer (RNN-T), attention mechanism, and non-autoregressive mask-predict models. Since each of these network architectures has pros and cons, a typical use case is to switch these separate models depending on the application requirement, resulting in the increased overhead of maintaining all models. Several methods for integrating two of these complementary models to mitigate the overhead issue have been proposed; however, if we integrate more models, we will further benefit from these complementary models and realize broader applications with a single system. This paper proposes four-decoder joint modeling (4D) of CTC, attention, RNN-T, and mask-predict, which has the following three advantages: 1) The four decoders are jointly trained so that they can be easily switched depending on the application scenarios. 2) Joint training may bring model regularization and improve the model robustness thanks to their complementary properties. 3) Novel one-pass joint decoding methods using CTC, attention, and RNN-T further improves the performance. The experimental results showed that the proposed model consistently reduced the WER.

Tenghao Huang, Donghee Lee, John Sweeney et al.

Flavor development in the food industry is increasingly challenged by the need for rapid innovation and precise flavor profile creation. Traditional flavor research methods typically rely on iterative, subjective testing, which lacks the efficiency and scalability required for modern demands. This paper presents three contributions to address the challenges. Firstly, we define a new problem domain for scientific agents in flavor science, conceptualized as the generation of hypotheses for flavor profile sourcing and understanding. To facilitate research in this area, we introduce the FoodPuzzle, a challenging benchmark consisting of 978 food items and 1,766 flavor molecules profiles. We propose a novel Scientific Agent approach, integrating in-context learning and retrieval augmented techniques to generate grounded hypotheses in the domain of food science. Experimental results indicate that our model significantly surpasses traditional methods in flavor profile prediction tasks, demonstrating its potential to transform flavor development practices.

Siddhant Arora, Jinchuan Tian, Jiatong Shi et al.

Reinforcement learning from human or AI feedback (RLHF/RLAIF) for speech-in/speech-out dialogue systems (SDS) remains underexplored, with prior work largely limited to single semantic rewards applied at the utterance level. Such setups overlook the multi-dimensional and multi-modal nature of conversational quality, which encompasses semantic coherence, audio naturalness, speaker consistency, emotion alignment, and turn-taking behavior. Moreover, they are fundamentally mismatched with duplex spoken dialogue systems that generate responses incrementally, where agents must make decisions based on partial utterances. We address these limitations with the first multi-reward RLAIF framework for SDS, combining semantic, audio-quality, and emotion-consistency rewards. To align utterance-level preferences with incremental, blockwise decoding in duplex models, we apply turn-level preference sampling and aggregate per-block log-probabilities within a single DPO objective. We present the first systematic study of preference learning for improving SDS quality in both multi-turn Chain-of-Thought and blockwise duplex models, and release a multi-reward DPO dataset to support reproducible research. Experiments show that single-reward RLAIF selectively improves its targeted metric, while joint multi-reward training yields consistent gains across semantic quality and audio naturalness. These results highlight the importance of holistic, multi-reward alignment for practical conversational SDS.

Takashi Maekaku, Jiatong Shi, Xuankai Chang et al.

Recently, the usefulness of self-supervised representation learning (SSRL) methods has been confirmed in various downstream tasks. Many of these models, as exemplified by HuBERT and WavLM, use pseudo-labels generated from spectral features or the model's own representation features. From previous studies, it is known that the pseudo-labels contain semantic information. However, the masked prediction task, the learning criterion of HuBERT, focuses on local contextual information and may not make effective use of global semantic information such as speaker, theme of speech, and so on. In this paper, we propose a new approach to enrich the semantic representation of HuBERT. We apply topic model to pseudo-labels to generate a topic label for each utterance. An auxiliary topic classification task is added to HuBERT by using topic labels as teachers. This allows additional global semantic information to be incorporated in an unsupervised manner. Experimental results demonstrate that our method achieves comparable or better performance than the baseline in most tasks, including automatic speech recognition and five out of the eight SUPERB tasks. Moreover, we find that topic labels include various information about utterance, such as gender, speaker, and its theme. This highlights the effectiveness of our approach in capturing multifaceted semantic nuances.

Yifan Peng, Jinchuan Tian, William Chen et al. · nvidia

Recent studies have highlighted the importance of fully open foundation models. The Open Whisper-style Speech Model (OWSM) is an initial step towards reproducing OpenAI Whisper using public data and open-source toolkits. However, previous versions of OWSM (v1 to v3) are still based on standard Transformer, which might lead to inferior performance compared to state-of-the-art speech encoder architectures. This work aims to improve the performance and efficiency of OWSM without additional data. We present a series of E-Branchformer-based models named OWSM v3.1, ranging from 100M to 1B parameters. OWSM v3.1 outperforms its predecessor, OWSM v3, in most evaluation benchmarks, while showing an improved inference speed of up to 25%. We further reveal the emergent ability of OWSM v3.1 in zero-shot contextual biasing speech recognition. We also provide a model trained on a subset of data with low license restrictions. We will publicly release the code, pre-trained models, and training logs.

Jinchuan Tian, Chunlei Zhang, Jiatong Shi et al.

Recent advancements in text-to-speech (TTS) have shown that language model (LM)-based systems offer competitive performance to their counterparts. Further optimization can be achieved through preference alignment algorithms, which adjust LMs to align with the preferences of reward models, enhancing the desirability of the generated content. This study presents a thorough empirical evaluation of how preference alignment algorithms, particularly Direct Preference Optimization (DPO), enhance LM-based TTS. With a 1.15B parameter LM-based TTS model, we demonstrate that preference alignment consistently improves intelligibility, speaker similarity, and proxy subjective evaluation scores, with the latter two metrics surpassing even human speech in certain evaluations. We also show preference alignment is applicable to low-resource scenarios and effectively generalized to out-of-domain applications.

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.

Jee-weon Jung, Wangyou Zhang, Jiatong Shi et al.

This paper introduces ESPnet-SPK, a toolkit designed with several objectives for training speaker embedding extractors. First, we provide an open-source platform for researchers in the speaker recognition community to effortlessly build models. We provide several models, ranging from x-vector to recent SKA-TDNN. Through the modularized architecture design, variants can be developed easily. We also aspire to bridge developed models with other domains, facilitating the broad research community to effortlessly incorporate state-of-the-art embedding extractors. Pre-trained embedding extractors can be accessed in an off-the-shelf manner and we demonstrate the toolkit's versatility by showcasing its integration with two tasks. Another goal is to integrate with diverse self-supervised learning features. We release a reproducible recipe that achieves an equal error rate of 0.39% on the Vox1-O evaluation protocol using WavLM-Large with ECAPA-TDNN.

Jinchuan Tian, Haoran Wang, Bo-Hao Su et al.

Current audio foundation models typically rely on rigid, task-specific supervision, addressing isolated factors of audio rather than the whole. In contrast, human intelligence processes audio holistically, seamlessly bridging physical signals with abstract cognitive concepts to execute complex tasks. Grounded in this philosophy, we introduce Bagpiper, an 8B audio foundation model that interprets physical audio via rich captions, i.e., comprehensive natural language descriptions that encapsulate the critical cognitive concepts inherent in the signal (e.g., transcription, audio events). By pre-training on a massive corpus of 600B tokens, the model establishes a robust bidirectional mapping between raw audio and this high-level conceptual space. During fine-tuning, Bagpiper adopts a caption-then-process workflow, simulating an intermediate cognitive reasoning step to solve diverse tasks without task-specific priors. Experimentally, Bagpiper outperforms Qwen-2.5-Omni on MMAU and AIRBench for audio understanding and surpasses CosyVoice3 and TangoFlux in generation quality, capable of synthesizing arbitrary compositions of speech, music, and sound effects. To the best of our knowledge, Bagpiper is among the first works that achieve unified understanding generation for general audio. Model, data, and code are available at Bagpiper Home Page.

Yichen Huang, Zachary Novack, Koichi Saito et al.

Despite significant recent advances in generative acoustic text-to-music (TTM) modeling, robust evaluation of these models lags behind, relying in particular on the popular Fréchet Audio Distance (FAD). In this work, we rigorously study the design space of reference-based divergence metrics for evaluating TTM models through (1) designing four synthetic meta-evaluations to measure sensitivity to particular musical desiderata, and (2) collecting and evaluating on MusicPrefs, the first open-source dataset of human preferences for TTM systems. We find that not only is the standard FAD setup inconsistent on both synthetic and human preference data, but that nearly all existing metrics fail to effectively capture desiderata, and are only weakly correlated with human perception. We propose a new metric, the MAUVE Audio Divergence (MAD), computed on representations from a self-supervised audio embedding model. We find that this metric effectively captures diverse musical desiderata (average rank correlation 0.84 for MAD vs. 0.49 for FAD and also correlates more strongly with MusicPrefs (0.62 vs. 0.14).

Jinchuan Tian, Jiatong Shi, William Chen et al. · nvidia

We present ESPnet-SpeechLM, an open toolkit designed to democratize the development of speech language models (SpeechLMs) and voice-driven agentic applications. The toolkit standardizes speech processing tasks by framing them as universal sequential modeling problems, encompassing a cohesive workflow of data preprocessing, pre-training, inference, and task evaluation. With ESPnet-SpeechLM, users can easily define task templates and configure key settings, enabling seamless and streamlined SpeechLM development. The toolkit ensures flexibility, efficiency, and scalability by offering highly configurable modules for every stage of the workflow. To illustrate its capabilities, we provide multiple use cases demonstrating how competitive SpeechLMs can be constructed with ESPnet-SpeechLM, including a 1.7B-parameter model pre-trained on both text and speech tasks, across diverse benchmarks. The toolkit and its recipes are fully transparent and reproducible at: https://github.com/espnet/espnet/tree/speechlm.

Jinchuan Tian, William Chen, Yifan Peng et al. · nvidia

This paper presents Open Unified Speech Language Models (OpusLMs), a family of open foundational speech language models (SpeechLMs) up to 7B. Initialized from decoder-only text language models, the OpusLMs are continuously pre-trained on 213K hours of speech-text pairs and 292B text-only tokens. We demonstrate our OpusLMs achieve comparable (or even superior) performance with existing SpeechLMs in speech recognition, speech synthesis, and text-only capabilities. Technically, this paper articulates our SpeechLM designs on tokenization, multi-stream language models, and multi-stage training strategies. We experimentally demonstrate the importance of model size scaling and the effect of annealing data selection. The OpusLMs are all built from publicly available materials and are fully transparent models. We release our code, data, checkpoints, and training logs to facilitate open SpeechLM research

Siddhant Arora, Yifan Peng, Jiatong Shi et al. · nvidia

Advancements in audio foundation models (FMs) have fueled interest in end-to-end (E2E) spoken dialogue systems, but different web interfaces for each system makes it challenging to compare and contrast them effectively. Motivated by this, we introduce an open-source, user-friendly toolkit designed to build unified web interfaces for various cascaded and E2E spoken dialogue systems. Our demo further provides users with the option to get on-the-fly automated evaluation metrics such as (1) latency, (2) ability to understand user input, (3) coherence, diversity, and relevance of system response, and (4) intelligibility and audio quality of system output. Using the evaluation metrics, we compare various cascaded and E2E spoken dialogue systems with a human-human conversation dataset as a proxy. Our analysis demonstrates that the toolkit allows researchers to effortlessly compare and contrast different technologies, providing valuable insights such as current E2E systems having poorer audio quality and less diverse responses. An example demo produced using our toolkit is publicly available here: https://huggingface.co/spaces/Siddhant/Voice_Assistant_Demo.

Jiatong Shi, Jionghao Han, Yichen Lu et al.

Role-play has become a key testbed for generative models, expanding from text-only dialogue to multimodal interaction. Extending role-play to speech captures prosody, emotion, and delivery, but also poses new evaluation challenges. Current pipelines often use audio large language models (ALLMs) as zero-shot judges, which miss paralinguistic cues, collapse multiple aspects into coarse scores, and rely on synthetic speech references that fail to reflect real-world roles. We present Speech-DRAME, a unified framework that contributes at three levels: (i) Speech-DRAME-EvalBench, an evaluation benchmark with bilingual human-annotated data and protocols for training and testing speech evaluation models (SEMs), (ii) DRAME-Eval, a fine-tuned evaluation model, which substantially outperforms zero-shot and few-shot ALLMs, and (iii) Speech-DRAME-RoleBench, a speech role-play benchmark that leverages DRAME-Eval as an automatic judge to compare speech foundation models (SFMs). Speech-DRAME distinguishes between two complementary evaluation strategies: Archetype Evaluation, a top-down approach measuring adherence to broad role archetypes, and Realism Evaluation, a bottom-up approach grounded in real human speech that emphasizes nuanced role quality. Compared to zero-shot ALLM judges, DRAME-Eval achieves stronger agreement with human ratings (Pearson correlation from 0.480 to 0.629 in archetypes, and 0.390 to 0.625 in realism). By integrating transparent benchmark resources, modeling approaches, and system-level evaluation, Speech-DRAME provides the first comprehensive, reproducible foundation for assessing spoken role-play.

Yuxun Tang, Lan Liu, Wenhao Feng et al.

Singing voice generation progresses rapidly, yet evaluating singing quality remains a critical challenge. Human subjective assessment, typically in the form of listening tests, is costly and time consuming, while existing objective metrics capture only limited perceptual aspects. In this work, we introduce SingMOS-Pro, a dataset for automatic singing quality assessment. Building on our preview version SingMOS, which provides only overall ratings, SingMOS-Pro expands annotations of the additional part to include lyrics, melody, and overall quality, offering broader coverage and greater diversity. The dataset contains 7,981 singing clips generated by 41 models across 12 datasets, spanning from early systems to recent advances. Each clip receives at least five ratings from professional annotators, ensuring reliability and consistency. Furthermore, we explore how to effectively utilize MOS data annotated under different standards and benchmark several widely used evaluation methods from related tasks on SingMOS-Pro, establishing strong baselines and practical references for future research. The dataset can be accessed at https://huggingface.co/datasets/TangRain/SingMOS-Pro.

Tomoki Hayashi, Ryuichi Yamamoto, Takenori Yoshimura et al.

This paper describes ESPnet2-TTS, an end-to-end text-to-speech (E2E-TTS) toolkit. ESPnet2-TTS extends our earlier version, ESPnet-TTS, by adding many new features, including: on-the-fly flexible pre-processing, joint training with neural vocoders, and state-of-the-art TTS models with extensions like full-band E2E text-to-waveform modeling, which simplify the training pipeline and further enhance TTS performance. The unified design of our recipes enables users to quickly reproduce state-of-the-art E2E-TTS results. We also provide many pre-trained models in a unified Python interface for inference, offering a quick means for users to generate baseline samples and build demos. Experimental evaluations with English and Japanese corpora demonstrate that our provided models synthesize utterances comparable to ground-truth ones, achieving state-of-the-art TTS performance. The toolkit is available online at https://github.com/espnet/espnet.

Jiatong Shi, Jonathan D. Amith, Rey Castillo García et al.

"Transcription bottlenecks", created by a shortage of effective human transcribers are one of the main challenges to endangered language (EL) documentation. Automatic speech recognition (ASR) has been suggested as a tool to overcome such bottlenecks. Following this suggestion, we investigated the effectiveness for EL documentation of end-to-end ASR, which unlike Hidden Markov Model ASR systems, eschews linguistic resources but is instead more dependent on large-data settings. We open source a Yoloxóchitl Mixtec EL corpus. First, we review our method in building an end-to-end ASR system in a way that would be reproducible by the ASR community. We then propose a novice transcription correction task and demonstrate how ASR systems and novice transcribers can work together to improve EL documentation. We believe this combinatory methodology would mitigate the transcription bottleneck and transcriber shortage that hinders EL documentation.

Pengcheng Guo, Florian Boyer, Xuankai Chang et al.

In this study, we present recent developments on ESPnet: End-to-End Speech Processing toolkit, which mainly involves a recently proposed architecture called Conformer, Convolution-augmented Transformer. This paper shows the results for a wide range of end-to-end speech processing applications, such as automatic speech recognition (ASR), speech translations (ST), speech separation (SS) and text-to-speech (TTS). Our experiments reveal various training tips and significant performance benefits obtained with the Conformer on different tasks. These results are competitive or even outperform the current state-of-art Transformer models. We are preparing to release all-in-one recipes using open source and publicly available corpora for all the above tasks with pre-trained models. Our aim for this work is to contribute to our research community by reducing the burden of preparing state-of-the-art research environments usually requiring high resources.

Jiatong Shi, Shuai Guo, Nan Huo et al.

The neural network (NN) based singing voice synthesis (SVS) systems require sufficient data to train well and are prone to over-fitting due to data scarcity. However, we often encounter data limitation problem in building SVS systems because of high data acquisition and annotation costs. In this work, we propose a Perceptual Entropy (PE) loss derived from a psycho-acoustic hearing model to regularize the network. With a one-hour open-source singing voice database, we explore the impact of the PE loss on various mainstream sequence-to-sequence models, including the RNN-based, transformer-based, and conformer-based models. Our experiments show that the PE loss can mitigate the over-fitting problem and significantly improve the synthesized singing quality reflected in objective and subjective evaluations.

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.

Ibrahim Said Ahmad, Antonios Anastasopoulos, Ondřej Bojar et al.

This paper reports on the shared tasks organized by the 21st IWSLT Conference. The shared tasks address 7 scientific challenges in spoken language translation: simultaneous and offline translation, automatic subtitling and dubbing, speech-to-speech translation, dialect and low-resource speech translation, and Indic languages. The shared tasks attracted 18 teams whose submissions are documented in 26 system papers. The growing interest towards spoken language translation is also witnessed by the constantly increasing number of shared task organizers and contributors to the overview paper, almost evenly distributed across industry and academia.

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/.

You Zhang, Yongyi Zang, Jiatong Shi et al.

The rapid advancement of AI-generated singing voices, which now closely mimic natural human singing and align seamlessly with musical scores, has led to heightened concerns for artists and the music industry. Unlike spoken voice, singing voice presents unique challenges due to its musical nature and the presence of strong background music, making singing voice deepfake detection (SVDD) a specialized field requiring focused attention. To promote SVDD research, we recently proposed the "SVDD Challenge," the very first research challenge focusing on SVDD for lab-controlled and in-the-wild bonafide and deepfake singing voice recordings. The challenge will be held in conjunction with the 2024 IEEE Spoken Language Technology Workshop (SLT 2024).

Siddhant Arora, Jinchuan Tian, Hayato Futami et al.

Unlike traditional cascaded pipelines, end-to-end (E2E) spoken dialogue systems preserve full differentiability and capture non-phonemic information, making them well-suited for modeling spoken interactions. However, existing E2E approaches often require large-scale training data and generates responses lacking semantic coherence. We propose a simple yet effective strategy leveraging a chain-of-thought (CoT) formulation, ensuring that training on conversational data remains closely aligned with the multimodal language model (LM)'s pre-training on speech recognition~(ASR), text-to-speech synthesis (TTS), and text LM tasks. Our method achieves over 1.5 ROUGE-1 improvement over the baseline, successfully training spoken dialogue systems on publicly available human-human conversation datasets, while being compute-efficient enough to train on just 300 hours of public human-human conversation data, such as the Switchboard. We will publicly release our models and training code.

Yi-Hui Chou, Kalvin Chang, Meng-Ju Wu et al.

Taiwanese Hokkien is declining in use and status due to a language shift towards Mandarin in Taiwan. This is partly why it is a low resource language in NLP and speech research today. To ensure that the state of the art in speech processing does not leave Taiwanese Hokkien behind, we contribute a 1.5-hour dataset of Taiwanese Hokkien to ML-SUPERB's hidden set. Evaluating ML-SUPERB's suite of self-supervised learning (SSL) speech representations on our dataset, we find that model size does not consistently determine performance. In fact, certain smaller models outperform larger ones. Furthermore, linguistic alignment between pretraining data and the target language plays a crucial role.

William Chen, Chutong Meng, Jiatong Shi et al.

Recent improvements in multilingual ASR have not been equally distributed across languages and language varieties. To advance state-of-the-art (SOTA) ASR models, we present the Interspeech 2025 ML-SUPERB 2.0 Challenge. We construct a new test suite that consists of data from 200+ languages, accents, and dialects to evaluate SOTA multilingual speech models. The challenge also introduces an online evaluation server based on DynaBench, allowing for flexibility in model design and architecture for participants. The challenge received 5 submissions from 3 teams, all of which outperformed our baselines. The best-performing submission achieved an absolute improvement in LID accuracy of 23% and a reduction in CER of 18% when compared to the best baseline on a general multilingual test set. On accented and dialectal data, the best submission obtained 30.2% lower CER and 15.7% higher LID accuracy, showing the importance of community challenges in making speech technologies more inclusive.

Kristin Qi, Jiatong Shi, Caroline Summerour et al.

Mild Cognitive Impairment (MCI) is an early stage of Alzheimer's disease (AD), a form of neurodegenerative disorder. Early identification of MCI is crucial for delaying its progression through timely interventions. Existing research has demonstrated the feasibility of detecting MCI using speech collected from clinical interviews or digital devices. However, these approaches typically analyze data collected at limited time points, limiting their ability to identify cognitive changes over time. This paper presents a longitudinal study using voice assistant systems (VAS) to remotely collect seven-session speech data at three-month intervals across 18 months. We propose two methods to improve MCI detection and the prediction of cognitive changes. The first method incorporates historical data, while the second predicts cognitive changes at two time points. Our results indicate improvements when incorporating historical data: the average F1-score for MCI detection improves from 58.6% to 71.2% (by 12.6%) in the case of acoustic features and from 62.1% to 75.1% (by 13.0%) in the case of linguistic features. Additionally, the prediction of cognitive changes achieves an F1-score of 73.7% in the case of acoustic features. These results confirm the potential of VAS-based speech sessions for early detection of cognitive decline.

Shih-Heng Wang, Zih-Ching Chen, Jiatong Shi et al. · meta-ai, mit

The utilization of speech Self-Supervised Learning (SSL) models achieves impressive performance on Automatic Speech Recognition (ASR). However, in low-resource language ASR, they encounter the domain mismatch problem between pre-trained and low-resource languages. Typical solutions like fine-tuning the SSL model suffer from high computation costs while using frozen SSL models as feature extractors comes with poor performance. To handle these issues, we extend a conventional efficient fine-tuning scheme based on the adapter. We add an extra intermediate adaptation to warm up the adapter and downstream model initialization. Remarkably, we update only 1-5% of the total model parameters to achieve the adaptation. Experimental results on the ML-SUPERB dataset show that our solution outperforms conventional efficient fine-tuning. It achieves up to a 28% relative improvement in the Character/Phoneme error rate when adapting to unseen languages.

Siddhant Arora, Jinchuan Tian, Hayato Futami et al.

Most end-to-end (E2E) spoken dialogue systems (SDS) rely on voice activity detection (VAD) for turn-taking, but VAD fails to distinguish between pauses and turn completions. Duplex SDS models address this by predicting output continuously, including silence tokens, thus removing the need for explicit VAD. However, they often have complex dual-channel architecture and lag behind cascaded models in semantic reasoning. To overcome these challenges, we propose SCoT: a Streaming Chain-of-Thought (CoT) framework for Duplex SDS, alternating between processing fixed-duration user input and generating responses in a blockwise manner. Using frame-level alignments, we create intermediate targets-aligned user transcripts and system responses for each block. Experiments show that our approach produces more coherent and interpretable responses than existing duplex methods while supporting lower-latency and overlapping interactions compared to turn-by-turn systems.