arXiv subjects

Jinyu Li

Publications and source records attributed to Jinyu Li.

At least 55 records · Page 3Linked to original sources

arXiv2025

AlignFormer: Modality Matching Can Achieve Better Zero-shot Instruction-Following Speech-LLM

Integrating speech into LLM (speech-LLM) has gaining increased attention recently. The mainstream solution is to connect a well-trained speech encoder and LLM with a neural adapter. However, the length mismatch between the speech and text sequences are not well handled, leading to imperfect modality matching between the speech and text. In this work, we propose a novel neural adapter, AlignFormer, to reduce the length gap between the two modalities. AlignFormer consists of CTC and dynamic-window QFormer layers, where the CTC alignment provides the dynamic window information for QFormer. The LLM backbone is frozen in training to preserve its text capability, especially the instruction following capability. When training with ASR data only, the proposed AlignFormer unlocks the instruction following capability for speech-LLM and the model can perform zero-shot speech translation (ST) and speech question answering (SQA) tasks. In fact, speech-LLM with AlignFormer can theoretically perform any tasks that the LLM backbone can deal with in the speech version. To evaluate the effectiveness of the instruction-following speech-LLM, we propose to use instruction following rate (IFR) and offer a systematic perspective for the IFR evaluation. In addition, we find that the audio position in training would affect the instruction following capability of speech-LLM and conduct an in-depth study on it. Our findings show that audio-first training achieves higher IFR than instruction-first training. The AlignFormer can achieve a near 100% IFR with audio-first training and game-changing improvements from zero to non-zero IFR on some evaluation data with instruction-first training. We believe that this study is a big step towards the perfect speech and text modality matching in the LLM embedding space.

eess.AS↗

arXiv2025

Improving Practical Aspects of End-to-End Multi-Talker Speech Recognition for Online and Offline Scenarios

We extend the frameworks of Serialized Output Training (SOT) to address practical needs of both streaming and offline automatic speech recognition (ASR) applications. Our approach focuses on balancing latency and accuracy, catering to real-time captioning and summarization requirements. We propose several key improvements: (1) Leveraging Continuous Speech Separation (CSS) single-channel front-end with end-to-end (E2E) systems for highly overlapping scenarios, challenging the conventional wisdom of E2E versus cascaded setups. The CSS framework improves the accuracy of the ASR system by separating overlapped speech from multiple speakers. (2) Implementing dual models -- Conformer Transducer for streaming and Sequence-to-Sequence for offline -- or alternatively, a two-pass model based on cascaded encoders. (3) Exploring segment-based SOT (segSOT) which is better suited for offline scenarios while also enhancing readability of multi-talker transcriptions.

eess.AS↗

arXiv2025

StreamMel: Real-Time Zero-shot Text-to-Speech via Interleaved Continuous Autoregressive Modeling

Recent advances in zero-shot text-to-speech (TTS) synthesis have achieved high-quality speech generation for unseen speakers, but most systems remain unsuitable for real-time applications because of their offline design. Current streaming TTS paradigms often rely on multi-stage pipelines and discrete representations, leading to increased computational cost and suboptimal system performance. In this work, we propose StreamMel, a pioneering single-stage streaming TTS framework that models continuous mel-spectrograms. By interleaving text tokens with acoustic frames, StreamMel enables low-latency, autoregressive synthesis while preserving high speaker similarity and naturalness. Experiments on LibriSpeech demonstrate that StreamMel outperforms existing streaming TTS baselines in both quality and latency. It even achieves performance comparable to offline systems while supporting efficient real-time generation, showcasing broad prospects for integration with real-time speech large language models. Audio samples are available at: https://aka.ms/StreamMel.

cs.SD↗

arXiv2025

PHRASED: Phrase Dictionary Biasing for Speech Translation

Phrases are essential to understand the core concepts in conversations. However, due to their rare occurrence in training data, correct translation of phrases is challenging in speech translation tasks. In this paper, we propose a phrase dictionary biasing method to leverage pairs of phrases mapping from the source language to the target language. We apply the phrase dictionary biasing method to two types of widely adopted models, a transducer-based streaming speech translation model and a multimodal large language model. Experimental results show that the phrase dictionary biasing method outperforms phrase list biasing by 21% relatively for the streaming speech translation model. In addition, phrase dictionary biasing enables multimodal large language models to use external phrase information, achieving 85% relative improvement in phrase recall.

cs.CL↗

arXiv2025

Length Aware Speech Translation for Video Dubbing

In video dubbing, aligning translated audio with the source audio is a significant challenge. Our focus is on achieving this efficiently, tailored for real-time, on-device video dubbing scenarios. We developed a phoneme-based end-to-end length-sensitive speech translation (LSST) model, which generates translations of varying lengths short, normal, and long using predefined tags. Additionally, we introduced length-aware beam search (LABS), an efficient approach to generate translations of different lengths in a single decoding pass. This approach maintained comparable BLEU scores compared to a baseline without length awareness while significantly enhancing synchronization quality between source and target audio, achieving a mean opinion score (MOS) gain of 0.34 for Spanish and 0.65 for Korean, respectively.

cs.CL↗

arXiv2025

V2SFlow: Video-to-Speech Generation with Speech Decomposition and Rectified Flow

In this paper, we introduce V2SFlow, a novel Video-to-Speech (V2S) framework designed to generate natural and intelligible speech directly from silent talking face videos. While recent V2S systems have shown promising results on constrained datasets with limited speakers and vocabularies, their performance often degrades on real-world, unconstrained datasets due to the inherent variability and complexity of speech signals. To address these challenges, we decompose the speech signal into manageable subspaces (content, pitch, and speaker information), each representing distinct speech attributes, and predict them directly from the visual input. To generate coherent and realistic speech from these predicted attributes, we employ a rectified flow matching decoder built on a Transformer architecture, which models efficient probabilistic pathways from random noise to the target speech distribution. Extensive experiments demonstrate that V2SFlow significantly outperforms state-of-the-art methods, even surpassing the naturalness of ground truth utterances. Code and models are available at: https://github.com/kaistmm/V2SFlow

cs.CV↗

arXiv2025

Autoregressive Speech Synthesis without Vector Quantization

We present MELLE, a novel continuous-valued token based language modeling approach for text-to-speech synthesis (TTS). MELLE autoregressively generates continuous mel-spectrogram frames directly from text condition, bypassing the need for vector quantization, which is typically designed for audio compression and sacrifices fidelity compared to continuous representations. Specifically, (i) instead of cross-entropy loss, we apply regression loss with a proposed spectrogram flux loss function to model the probability distribution of the continuous-valued tokens; (ii) we have incorporated variational inference into MELLE to facilitate sampling mechanisms, thereby enhancing the output diversity and model robustness. Experiments demonstrate that, compared to the two-stage codec language model VALL-E and its variants, the single-stage MELLE mitigates robustness issues by avoiding the inherent flaws of sampling vector-quantized codes, achieves superior performance across multiple metrics, and, most importantly, offers a more streamlined paradigm. The demos of our work are provided at https://aka.ms/melle.

cs.CL↗

arXiv2025

TS3-Codec: Transformer-Based Simple Streaming Single Codec

Neural audio codecs (NACs) have garnered significant attention as key technologies for audio compression as well as audio representation for speech language models. While mainstream NAC models are predominantly convolution-based, the performance of NACs with a purely transformer-based, and convolution-free architecture remains unexplored. This paper introduces TS3-Codec, a Transformer-Based Simple Streaming Single Codec. TS3-Codec consists of only a stack of transformer layers with a few linear layers, offering greater simplicity and expressiveness by fully eliminating convolution layers that require careful hyperparameter tuning and large computations. Under the streaming setup, the proposed TS3-Codec achieves comparable or superior performance compared to the codec with state-of-the-art convolution-based architecture while requiring only 12% of the computation and 77% of bitrate. Furthermore, it significantly outperforms the convolution-based codec when using similar computational resources.

eess.AS↗

arXiv2025

ARLON: Boosting Diffusion Transformers with Autoregressive Models for Long Video Generation

Text-to-video models have recently undergone rapid and substantial advancements. Nevertheless, due to limitations in data and computational resources, achieving efficient generation of long videos with rich motion dynamics remains a significant challenge. To generate high-quality, dynamic, and temporally consistent long videos, this paper presents ARLON, a novel framework that boosts diffusion Transformers with autoregressive models for long video generation, by integrating the coarse spatial and long-range temporal information provided by the AR model to guide the DiT model. Specifically, ARLON incorporates several key innovations: 1) A latent Vector Quantized Variational Autoencoder (VQ-VAE) compresses the input latent space of the DiT model into compact visual tokens, bridging the AR and DiT models and balancing the learning complexity and information density; 2) An adaptive norm-based semantic injection module integrates the coarse discrete visual units from the AR model into the DiT model, ensuring effective guidance during video generation; 3) To enhance the tolerance capability of noise introduced from the AR inference, the DiT model is trained with coarser visual latent tokens incorporated with an uncertainty sampling module. Experimental results demonstrate that ARLON significantly outperforms the baseline OpenSora-V1.2 on eight out of eleven metrics selected from VBench, with notable improvements in dynamic degree and aesthetic quality, while delivering competitive results on the remaining three and simultaneously accelerating the generation process. In addition, ARLON achieves state-of-the-art performance in long video generation. Detailed analyses of the improvements in inference efficiency are presented, alongside a practical application that demonstrates the generation of long videos using progressive text prompts. See demos of ARLON at http://aka.ms/arlon.

cs.CV↗

arXiv2025

Predicting Accurate X-ray Absorption Spectra for CN$^+$, CN, and CN$^-$: Insights from Multiconfigurational and Density Functional Simulations

High-resolution X-ray spectroscopy is an essential tool in X-ray astronomy, enabling detailed studies of celestial objects and their physical and chemical properties. However, comprehensive mapping of high-resolution X-ray spectra for even simple interstellar and circumstellar molecules is still lacking. In this study, we conducted systematic quantum chemical simulations to predict the C1s X-ray absorption spectra of CN$^+$, CN, and CN$^-$. Our findings provide valuable references for both X-ray astronomy and laboratory studies. We assigned the first electronic peak of CN$^+$ and CN to C1s $\rightarrow σ^*$ transitions, while the peak for CN$^-$ corresponds to a C1s $\rightarrow π^*$ transition. We explained that the two-fold degeneracy ($π^*_{xz}$ and $π^*_{yz}$) of the C1s$\rightarrowπ^*$ transitions results in CN$^-$ exhibiting a significantly stronger first absorption compared to the other two systems. We further calculated the vibronic fine structures for these transitions using the quantum wavepacket method based on multiconfigurational-level, anharmonic potential energy curves, revealing distinct energy positions for the 0-0 absorptions at 280.7 eV, 279.6 eV, and 285.8 eV. Each vibronic profile features a prominent 0-0 peak, showing overall similarity but differing intensity ratios of the 0-0 and 0-1 peaks. Notably, introducing a C1s core hole leads to shortened C-N bond lengths and increased vibrational frequencies across all species. These findings enhance our understanding of the electronic structures and X-ray spectra of carbon-nitrogen species, emphasizing the influence of charge state on X-ray absorptions.

physics.chem-ph↗

arXiv2025

Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAs

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.

cs.CL↗

arXiv2025

Sce2DriveX: A Generalized MLLM Framework for Scene-to-Drive Learning

End-to-end autonomous driving, which directly maps raw sensor inputs to low-level vehicle controls, is an important part of Embodied AI. Despite successes in applying Multimodal Large Language Models (MLLMs) for high-level traffic scene semantic understanding, it remains challenging to effectively translate these conceptual semantics understandings into low-level motion control commands and achieve generalization and consensus in cross-scene driving. We introduce Sce2DriveX, a human-like driving chain-of-thought (CoT) reasoning MLLM framework. Sce2DriveX utilizes multimodal joint learning from local scene videos and global BEV maps to deeply understand long-range spatiotemporal relationships and road topology, enhancing its comprehensive perception and reasoning capabilities in 3D dynamic/static scenes and achieving driving generalization across scenes. Building on this, it reconstructs the implicit cognitive chain inherent in human driving, covering scene understanding, meta-action reasoning, behavior interpretation analysis, motion planning and control, thereby further bridging the gap between autonomous driving and human thought processes. To elevate model performance, we have developed the first extensive Visual Question Answering (VQA) driving instruction dataset tailored for 3D spatial understanding and long-axis task reasoning. Extensive experiments demonstrate that Sce2DriveX achieves state-of-the-art performance from scene understanding to end-to-end driving, as well as robust generalization on the CARLA Bench2Drive benchmark.

cs.CV↗

arXiv2025

Streaming Speaker Change Detection and Gender Classification for Transducer-Based Multi-Talker Speech Translation

Streaming multi-talker speech translation is a task that involves not only generating accurate and fluent translations with low latency but also recognizing when a speaker change occurs and what the speaker's gender is. Speaker change information can be used to create audio prompts for a zero-shot text-to-speech system, and gender can help to select speaker profiles in a conventional text-to-speech model. We propose to tackle streaming speaker change detection and gender classification by incorporating speaker embeddings into a transducer-based streaming end-to-end speech translation model. Our experiments demonstrate that the proposed methods can achieve high accuracy for both speaker change detection and gender classification.

cs.SD↗

arXiv2025

Addressing speaker gender bias in large scale speech translation systems

This study addresses the issue of speaker gender bias in Speech Translation (ST) systems, which can lead to offensive and inaccurate translations. The masculine bias often found in large-scale ST systems is typically perpetuated through training data derived from Machine Translation (MT) systems. Our approach involves two key steps. First, we employ Large Language Models (LLMs) to rectify translations based on the speaker's gender in a cost-effective manner. Second, we fine-tune the ST model with the corrected data, enabling the model to generate gender-specific translations directly from audio cues, without the need for explicit gender input. Additionally, we propose a three-mode fine-tuned model for scenarios where the speaker's gender is either predefined or should not be inferred from speech cues. We demonstrate a 70% improvement in translations for female speakers compared to our baseline and other large-scale ST systems, such as Seamless M4T and Canary, on the MuST-SHE test set.

cs.CL↗

arXiv2024

SLAM-Omni: Timbre-Controllable Voice Interaction System with Single-Stage Training

Recent advancements highlight the potential of end-to-end real-time spoken dialogue systems, showcasing their low latency and high quality. In this paper, we introduce SLAM-Omni, a timbre-controllable, end-to-end voice interaction system with single-stage training. SLAM-Omni achieves zero-shot timbre control by modeling spoken language with semantic tokens and decoupling speaker information to a vocoder. By predicting grouped speech semantic tokens at each step, our method significantly reduces the sequence length of audio tokens, accelerating both training and inference. Additionally, we propose historical text prompting to compress dialogue history, facilitating efficient multi-round interactions. Comprehensive evaluations reveal that SLAM-Omni outperforms prior models of similar scale, requiring only 15 hours of training on 4 GPUs with limited data. Notably, it is the first spoken dialogue system to achieve competitive performance with a single-stage training approach, eliminating the need for pre-training on TTS or ASR tasks. Further experiments validate its multilingual and multi-turn dialogue capabilities on larger datasets.

eess.AS↗

arXiv2024

CoVoMix: Advancing Zero-Shot Speech Generation for Human-like Multi-talker Conversations

Recent advancements in zero-shot text-to-speech (TTS) modeling have led to significant strides in generating high-fidelity and diverse speech. However, dialogue generation, along with achieving human-like naturalness in speech, continues to be a challenge. In this paper, we introduce CoVoMix: Conversational Voice Mixture Generation, a novel model for zero-shot, human-like, multi-speaker, multi-round dialogue speech generation. CoVoMix first converts dialogue text into multiple streams of discrete tokens, with each token stream representing semantic information for individual talkers. These token streams are then fed into a flow-matching based acoustic model to generate mixed mel-spectrograms. Finally, the speech waveforms are produced using a HiFi-GAN model. Furthermore, we devise a comprehensive set of metrics for measuring the effectiveness of dialogue modeling and generation. Our experimental results show that CoVoMix can generate dialogues that are not only human-like in their naturalness and coherence but also involve multiple talkers engaging in multiple rounds of conversation. This is exemplified by instances generated in a single channel where one speaker's utterance is seamlessly mixed with another's interjections or laughter, indicating the latter's role as an attentive listener. Audio samples are available at https://aka.ms/covomix.

eess.AS↗

arXiv2024

TransVIP: Speech to Speech Translation System with Voice and Isochrony Preservation

There is a rising interest and trend in research towards directly translating speech from one language to another, known as end-to-end speech-to-speech translation. However, most end-to-end models struggle to outperform cascade models, i.e., a pipeline framework by concatenating speech recognition, machine translation and text-to-speech models. The primary challenges stem from the inherent complexities involved in direct translation tasks and the scarcity of data. In this study, we introduce a novel model framework TransVIP that leverages diverse datasets in a cascade fashion yet facilitates end-to-end inference through joint probability. Furthermore, we propose two separated encoders to preserve the speaker's voice characteristics and isochrony from the source speech during the translation process, making it highly suitable for scenarios such as video dubbing. Our experiments on the French-English language pair demonstrate that our model outperforms the current state-of-the-art speech-to-speech translation model.

cs.CL↗

arXiv2024

Failing Forward: Improving Generative Error Correction for ASR with Synthetic Data and Retrieval Augmentation

Generative Error Correction (GEC) has emerged as a powerful post-processing method to enhance the performance of Automatic Speech Recognition (ASR) systems. However, we show that GEC models struggle to generalize beyond the specific types of errors encountered during training, limiting their ability to correct new, unseen errors at test time, particularly in out-of-domain (OOD) scenarios. This phenomenon amplifies with named entities (NEs), where, in addition to insufficient contextual information or knowledge about the NEs, novel NEs keep emerging. To address these issues, we propose DARAG (Data- and Retrieval-Augmented Generative Error Correction), a novel approach designed to improve GEC for ASR in in-domain (ID) and OOD scenarios. We augment the GEC training dataset with synthetic data generated by prompting LLMs and text-to-speech models, thereby simulating additional errors from which the model can learn. For OOD scenarios, we simulate test-time errors from new domains similarly and in an unsupervised fashion. Additionally, to better handle named entities, we introduce retrieval-augmented correction by augmenting the input with entities retrieved from a database. Our approach is simple, scalable, and both domain- and language-agnostic. We experiment on multiple datasets and settings, showing that DARAG outperforms all our baselines, achieving 8\% -- 30\% relative WER improvements in ID and 10\% -- 33\% improvements in OOD settings.

eess.AS↗