arXiv subjects

Kuan-Po Huang

Publications and source records attributed to Kuan-Po Huang.

At least 19 recordsLinked to original sources

arXiv2026

Fréchet Distance Loss on Speech Representations for Text-to-Speech Synthesis

Few-step diffusion and flow-matching text-to-speech (TTS) models are usually trained with local objectives, such as conditional flow matching, reconstruction, and stop prediction. These losses provide stable optimization, but they never ask whether sampled speech follows the distribution of high-quality speech. We propose Speech Representation Fr'echet Distance loss (SR-FD), a training-time distributional regularizer for tokenizer-free flow-matching autoregressive TTS. During fine-tuning, the model synthesizes speech with the same few-step sampler used at deployment, and SR-FD matches the mean and covariance of frozen Whisper and CTC features of this speech to reference statistics computed offline from three complementary content targets. The loss requires no discriminator and no inference-time computation. On Seed-TTS English, four-step SR-FD fine-tuning reduces WER from the original four-step VoxCPM2 baseline's 2.2279% to 1.4147%, a 36.5% relative reduction, and also surpasses the original ten-step baseline at 1.7366%; both gains are significant under an utterance-level paired bootstrap. Speaker similarity and objective quality proxies are preserved at the ten-step level, and an error analysis shows the gain comes from content substitutions across all prompt lengths. SR-FD is thus an intelligibility-improving distributional regularizer for few-step TTS.

cs.SD↗

arXiv2026

FdAudio: MeanFlow-Anchored Fréchet-Distance Post-Training for One-Step Text-to-Audio Generation

While recent few-step sampling text-to-audio generation models like MeanAudio substantially accelerate generation by modeling average velocities, their strict one-step generation quality still lags significantly behind multi-step counterparts. We propose FdAudio to bridge this gap. Unlike MeanAudio, which relies solely on regression against target velocity fields, our post-training approach optimizes the final one-step distribution directly across pre-trained embedding spaces via a multi-representation Fréchet-distance (FD) loss. Crucially, to prevent the multi-step degradation that naive post-training with FD-loss causes, we introduce a MeanFlow consistency objective as a structural anchor. Results demonstrate that FdAudio establishes state-of-the-art one-step T2A generation quality among few-step systems, yielding an 11.4% reduction in FD score and a 28.8% improvement in FAD score relative to the baseline MeanAudio framework. Notably, we solve FD post-training's naive multi-step degradation issue by proposing the MeanFlow anchor, enabling a 25-step sampling path to maintain high-fidelity audio synthesis that matches or surpasses strong multi-step models at a fraction of their computational latency.

eess.AS↗

arXiv2026

Fast Text-to-Audio Generation with One-Step Sampling via Energy-Scoring and Auxiliary Contextual Representation Distillation

Autoregressive (AR) models with diffusion heads have recently achieved strong text-to-audio performance, yet their iterative decoding and multi-step sampling process introduce high-latency issues. To address this bottleneck, we propose a one-step sampling framework that combines an energy-distance training objective with representation-level distillation. An energy-scoring head maps Gaussian noise directly to audio latents in one step, eliminating the need for a costly recursive diffusion sampling process, while distillation from a masked autoregressive (MAR) text-to-audio model preserves the strong conditioning learned during diffusion training. On the AudioCaps benchmark, our method consistently outperforms prior one-step baselines such as ConsistencyTTA, SoundCTM, AudioLCM and AudioTurbo, on both objective and subjective metrics, while substantially narrowing the quality gap to AR diffusion systems with multi-step sampling. Compared to the state-of-the-art AR diffusion system, IMPACT, our approach achieves up to $8.5$x faster batch inference with highly competitive audio quality. These results demonstrate that combining energy-distance training with representation-level distillation provides an effective recipe for fast, high-quality text-to-audio synthesis.

cs.SD↗

arXiv2026

DeSTA2.5-Audio: Toward General-Purpose Large Audio Language Model with Self-Generated Cross-Modal Alignment

We introduce DeSTA2.5-Audio, a general-purpose Large Audio Language Model (LALM) designed for robust auditory perception and instruction-following. Recent LALMs augment Large Language Models (LLMs) with auditory capabilities by training on large-scale audio-instruction datasets. However, existing LALMs have often suffered from the catastrophic forgetting of the LLM's original abilities. Therefore, balancing knowledge retention and audio perception has become a critical challenge. To address this, we revisit the data construction pipeline and propose a self-generated cross-modal alignment strategy in which the backbone LLM generates its own training targets, named DeSTA. This approach aims at preserving the LLM's native language proficiency thereby enabling zero-shot generalization without task-specific tuning. We construct DeSTA-AQA5M, a large-scale, task-agnostic dataset containing 5 million training samples derived from 7,000 hours of audio spanning 50 diverse datasets, including speech, environmental sounds, and music. DeSTA2.5-Audio achieves state-of-the-art or competitive performance across a wide range of audio-language benchmarks, including Dynamic-SUPERB, MMAU, SAKURA, Speech-IFEval, and VoiceBench. Comprehensive comparative studies demonstrate that our self-generated strategy outperforms existing training strategies. Our findings underscore the importance of carefully designed data construction in LALM development and offer practical insights for building robust, general-purpose LALMs.

eess.AS↗

arXiv2025

Generative Audio Language Modeling with Continuous-valued Tokens and Masked Next-Token Prediction

Autoregressive next-token prediction with the Transformer decoder has become a de facto standard in large language models (LLMs), achieving remarkable success in Natural Language Processing (NLP) at scale. Extending this paradigm to audio poses unique challenges due to its inherently continuous nature. We research audio generation with a causal language model (LM) without discrete tokens. We leverage token-wise diffusion to model the continuous distribution of the next continuous-valued token. Our approach delivers significant improvements over previous discrete solution, AudioGen, achieving 20% and 40% relative gains on AudioCaps in Frechet Audio Distance (FAD) and Kullback-Leibler (KL) divergence, respectively. Additionally, we propose a novel masked next-token prediction task that incorporates masked prediction into the causal LM framework. On AudioCaps, the innovation yields 41% and 33% relative FAD improvements over AudioGen Base (285M) and AudioGen Large (1B) models, respectively, and is on par with the state-of-the-art (SOTA) diffusion models. Furthermore, we achieve these results with significantly fewer parameters -- 193M for our Base and 462M for our Large models.

eess.AS↗

arXiv2025

Dynamic-SUPERB Phase-2: A Collaboratively Expanding Benchmark for Measuring the Capabilities of Spoken Language Models with 180 Tasks

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.

cs.CL↗

arXiv2025

IMPACT: Iterative Mask-based Parallel Decoding for Text-to-Audio Generation with Diffusion Modeling

Text-to-audio generation synthesizes realistic sounds or music given a natural language prompt. Diffusion-based frameworks, including the Tango and the AudioLDM series, represent the state-of-the-art in text-to-audio generation. Despite achieving high audio fidelity, they incur significant inference latency due to the slow diffusion sampling process. MAGNET, a mask-based model operating on discrete tokens, addresses slow inference through iterative mask-based parallel decoding. However, its audio quality still lags behind that of diffusion-based models. In this work, we introduce IMPACT, a text-to-audio generation framework that achieves high performance in audio quality and fidelity while ensuring fast inference. IMPACT utilizes iterative mask-based parallel decoding in a continuous latent space powered by diffusion modeling. This approach eliminates the fidelity constraints of discrete tokens while maintaining competitive inference speed. Results on AudioCaps demonstrate that IMPACT achieves state-of-the-art performance on key metrics including Fréchet Distance (FD) and Fréchet Audio Distance (FAD) while significantly reducing latency compared to prior models. The project website is available at https://audio-impact.github.io/.

eess.AS↗

arXiv2025

How to Learn a New Language? An Efficient Solution for Self-Supervised Learning Models Unseen Languages Adaption in Low-Resource Scenario

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.

cs.SD↗

arXiv2024

Enhancing Multilingual ASR for Unseen Languages via Language Embedding Modeling

Multilingual Automatic Speech Recognition (ASR) aims to recognize and transcribe speech from multiple languages within a single system. Whisper, one of the most advanced ASR models, excels in this domain by handling 99 languages effectively, leveraging a vast amount of data and incorporating language tags as prefixes to guide the recognition process. However, despite its success, Whisper struggles with unseen languages, those not included in its pre-training. Motivated by the observation that many languages share linguistic characteristics, we propose methods that exploit these relationships to enhance ASR performance on unseen languages. Specifically, we introduce a weighted sum method, which computes a weighted sum of the embeddings of language tags, using Whisper's predicted language probabilities. In addition, we develop a predictor-based approach that refines the weighted sum embedding to more closely approximate the true embedding for unseen languages. Experimental results demonstrate substantial improvements in ASR performance, both in zero-shot and fine-tuning settings. Our proposed methods outperform baseline approaches, providing an effective solution for addressing unseen languages in multilingual ASR.

eess.AS↗

arXiv2024

Do Prompts Really Prompt? Exploring the Prompt Understanding Capability of Whisper

This research explores how the information of prompts interacts with the high-performing speech recognition model, Whisper. We compare its performances when prompted by prompts with correct information and those corrupted with incorrect information. Our results unexpectedly show that Whisper may not understand the textual prompts in a human-expected way. Additionally, we find that performance improvement is not guaranteed even with stronger adherence to the topic information in textual prompts. It is also noted that English prompts generally outperform Mandarin ones on datasets of both languages, likely due to differences in training data distributions for these languages despite the mismatch with pre-training scenarios. Conversely, we discover that Whisper exhibits awareness of misleading information in language tokens by ignoring incorrect language tokens and focusing on the correct ones. In sum, We raise insightful questions about Whisper's prompt understanding and reveal its counter-intuitive behaviors. We encourage further studies.

cs.CL↗

arXiv2024

Dataset-Distillation Generative Model for Speech Emotion Recognition

Deep learning models for speech rely on large datasets, presenting computational challenges. Yet, performance hinges on training data size. Dataset Distillation (DD) aims to learn a smaller dataset without much performance degradation when training with it. DD has been investigated in computer vision but not yet in speech. This paper presents the first approach for DD to speech targeting Speech Emotion Recognition on IEMOCAP. We employ Generative Adversarial Networks (GANs) not to mimic real data but to distil key discriminative information of IEMOCAP that is useful for downstream training. The GAN then replaces the original dataset and can sample custom synthetic dataset sizes. It performs comparably when following the original class imbalance but improves performance by 0.3% absolute UAR with balanced classes. It also reduces dataset storage and accelerates downstream training by 95% in both cases and reduces speaker information which could help for a privacy application.

cs.SD↗

arXiv2024

Zero Resource Code-switched Speech Benchmark Using Speech Utterance Pairs For Multiple Spoken Languages

We introduce a new zero resource code-switched speech benchmark designed to directly assess the code-switching capabilities of self-supervised speech encoders. We showcase a baseline system of language modeling on discrete units to demonstrate how the code-switching abilities of speech encoders can be assessed in a zero-resource manner. Our experiments encompass a variety of well-known speech encoders, including Wav2vec 2.0, HuBERT, XLSR, etc. We examine the impact of pre-training languages and model size on benchmark performance. Notably, though our results demonstrate that speech encoders with multilingual pre-training, exemplified by XLSR, outperform monolingual variants (Wav2vec 2.0, HuBERT) in code-switching scenarios, there is still substantial room for improvement in their code-switching linguistic abilities.

eess.AS↗

arXiv2023

Investigating Zero-Shot Generalizability on Mandarin-English Code-Switched ASR and Speech-to-text Translation of Recent Foundation Models with Self-Supervision and Weak Supervision

This work evaluated several cutting-edge large-scale foundation models based on self-supervision or weak supervision, including SeamlessM4T, SeamlessM4T v2, and Whisper-large-v3, on three code-switched corpora. We found that self-supervised models can achieve performances close to the supervised model, indicating the effectiveness of multilingual self-supervised pre-training. We also observed that these models still have room for improvement as they kept making similar mistakes and had unsatisfactory performances on modeling intra-sentential code-switching. In addition, the validity of several variants of Whisper was explored, and we concluded that they remained effective in a code-switching scenario, and similar techniques for self-supervised models are worth studying to boost the performance of code-switched tasks.

eess.AS↗

arXiv2023

Noise robust distillation of self-supervised speech models via correlation metrics

Compared to large speech foundation models, small distilled models exhibit degraded noise robustness. The student's robustness can be improved by introducing noise at the inputs during pre-training. Despite this, using the standard distillation loss still yields a student with degraded performance. Thus, this paper proposes improving student robustness via distillation with correlation metrics. Teacher behavior is learned by maximizing the teacher and student cross-correlation matrix between their representations towards identity. Noise robustness is encouraged via the student's self-correlation minimization. The proposed method is agnostic of the teacher model and consistently outperforms the previous approach. This work also proposes an heuristic to weigh the importance of the two correlation terms automatically. Experiments show consistently better clean and noise generalization on Intent Classification, Keyword Spotting, and Automatic Speech Recognition tasks on SUPERB Challenge.

cs.SD↗

arXiv2023

Ensemble knowledge distillation of self-supervised speech models

Distilled self-supervised models have shown competitive performance and efficiency in recent years. However, there is a lack of experience in jointly distilling multiple self-supervised speech models. In our work, we performed Ensemble Knowledge Distillation (EKD) on various self-supervised speech models such as HuBERT, RobustHuBERT, and WavLM. We tried two different aggregation techniques, layerwise-average and layerwise-concatenation, to the representations of different teacher models and found that the former was more effective. On top of that, we proposed a multiple prediction head method for student models to predict different layer outputs of multiple teacher models simultaneously. The experimental results show that our method improves the performance of the distilled models on four downstream speech processing tasks, Phoneme Recognition, Speaker Identification, Emotion Recognition, and Automatic Speech Recognition in the hidden-set track of the SUPERB benchmark.

eess.AS↗

arXiv2022

Improving generalizability of distilled self-supervised speech processing models under distorted settings

Self-supervised learned (SSL) speech pre-trained models perform well across various speech processing tasks. Distilled versions of SSL models have been developed to match the needs of on-device speech applications. Though having similar performance as original SSL models, distilled counterparts suffer from performance degradation even more than their original versions in distorted environments. This paper proposes to apply Cross-Distortion Mapping and Domain Adversarial Training to SSL models during knowledge distillation to alleviate the performance gap caused by the domain mismatch problem. Results show consistent performance improvements under both in- and out-of-domain distorted setups for different downstream tasks while keeping efficient model size.

cs.SD↗

arXiv2022

Improving the transferability of speech separation by meta-learning

Speech separation aims to separate multiple speech sources from a speech mixture. Although speech separation is well-solved on some existing English speech separation benchmarks, it is worthy of more investigation on the generalizability of speech separation models on the accents or languages unseen during training. This paper adopts meta-learning based methods to improve the transferability of speech separation models. With the meta-learning based methods, we discovered that only using speech data with one accent, the native English accent, as our training data, the models still can be adapted to new unseen accents on the Speech Accent Archive. We compared the results with a human-rated native-likeness of accents, showing that the transferability of MAML methods has less relation to the similarity of data between the training and testing phase compared to the typical transfer learning methods. Furthermore, we found that models can deal with different language data from the CommonVoice corpus during the testing phase. Most of all, the MAML methods outperform typical transfer learning methods when it comes to new accents, new speakers, new languages, and noisy environments.

cs.SD↗

arXiv2021

Multi-accent Speech Separation with One Shot Learning

Speech separation is a problem in the field of speech processing that has been studied in full swing recently. However, there has not been much work studying a multi-accent speech separation scenario. Unseen speakers with new accents and noise aroused the domain mismatch problem which cannot be easily solved by conventional joint training methods. Thus, we applied MAML and FOMAML to tackle this problem and obtained higher average Si-SNRi values than joint training on almost all the unseen accents. This proved that these two methods do have the ability to generate well-trained parameters for adapting to speech mixtures of new speakers and accents. Furthermore, we found out that FOMAML obtains similar performance compared to MAML while saving a lot of time.

cs.SD↗