Yongyi Zang

Yongyi Zang, Yixiao Zhang · bytedance

Large-scale text-to-music generation models have significantly enhanced music creation capabilities, offering unprecedented creative freedom. However, their ability to collaborate effectively with human musicians remains limited. In this paper, we propose a framework to describe the musical interaction process, which includes expression, interpretation, and execution of controls. Following this framework, we argue that the primary gap between existing text-to-music models and musicians lies in the interpretation stage, where models lack the ability to interpret controls from musicians. We also propose two strategies to address this gap and call on the music information retrieval community to tackle the interpretation challenge to improve human-AI musical collaboration.

Yongyi Zang, You Zhang, Mojtaba Heydari et al.

The rise of singing voice synthesis presents critical challenges to artists and industry stakeholders over unauthorized voice usage. Unlike synthesized speech, synthesized singing voices are typically released in songs containing strong background music that may hide synthesis artifacts. Additionally, singing voices present different acoustic and linguistic characteristics from speech utterances. These unique properties make singing voice deepfake detection a relevant but significantly different problem from synthetic speech detection. In this work, we propose the singing voice deepfake detection task. We first present SingFake, the first curated in-the-wild dataset consisting of 28.93 hours of bonafide and 29.40 hours of deepfake song clips in five languages from 40 singers. We provide a train/validation/test split where the test sets include various scenarios. We then use SingFake to evaluate four state-of-the-art speech countermeasure systems trained on speech utterances. We find these systems lag significantly behind their performance on speech test data. When trained on SingFake, either using separated vocal tracks or song mixtures, these systems show substantial improvement. However, our evaluations also identify challenges associated with unseen singers, communication codecs, languages, and musical contexts, calling for dedicated research into singing voice deepfake detection. The SingFake dataset and related resources are available at https://www.singfake.org/.

Ruibin Yuan, Hanfeng Lin, Shuyue Guo et al.

We tackle the task of long-form music generation--particularly the challenging \textbf{lyrics-to-song} problem--by introducing YuE, a family of open foundation models based on the LLaMA2 architecture. Specifically, YuE scales to trillions of tokens and generates up to five minutes of music while maintaining lyrical alignment, coherent musical structure, and engaging vocal melodies with appropriate accompaniment. It achieves this through (1) track-decoupled next-token prediction to overcome dense mixture signals, (2) structural progressive conditioning for long-context lyrical alignment, and (3) a multitask, multiphase pre-training recipe to converge and generalize. In addition, we redesign the in-context learning technique for music generation, enabling versatile style transfer (e.g., converting Japanese city pop into an English rap while preserving the original accompaniment) and bidirectional generation. Through extensive evaluation, we demonstrate that YuE matches or even surpasses some of the proprietary systems in musicality and vocal agility. In addition, fine-tuning YuE enables additional controls and enhanced support for tail languages. Furthermore, beyond generation, we show that YuE's learned representations can perform well on music understanding tasks, where the results of YuE match or exceed state-of-the-art methods on the MARBLE benchmark. Keywords: lyrics2song, song generation, long-form, foundation model, music generation

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).

Dichucheng Li, Yongyi Zang, Qiuqiang Kong

Automatic Music Transcription (AMT), aiming to get musical notes from raw audio, typically uses frame-level systems with piano-roll outputs or language model (LM)-based systems with note-level predictions. However, frame-level systems require manual thresholding, while the LM-based systems struggle with long sequences. In this paper, we propose a hybrid method combining pre-trained roll-based encoders with an LM decoder to leverage the strengths of both methods. Besides, our approach employs a hierarchical prediction strategy, first predicting onset and pitch, then velocity, and finally offset. The hierarchical prediction strategy reduces computational costs by breaking down long sequences into different hierarchies. Evaluated on two benchmark roll-based encoders, our method outperforms traditional piano-roll outputs 0.01 and 0.022 in onset-offset-velocity F1 score, demonstrating its potential as a performance-enhancing plug-in for arbitrary roll-based music transcription encoder.

Yongyi Zang, Jiarui Hai, Wanying Ge et al.

Music Source Restoration (MSR) aims to recover original, unprocessed instrument stems from professionally mixed and degraded audio, requiring the reversal of both production effects and real-world degradations. We present the inaugural MSR Challenge, which features objective evaluation on studio-produced mixtures using Multi-Mel-SNR, Zimtohrli, and FAD-CLAP, alongside subjective evaluation on real-world degraded recordings. Five teams participated in the challenge. The winning system achieved 4.46 dB Multi-Mel-SNR and 3.47 MOS-Overall, corresponding to relative improvements of 91% and 18% over the second-place system, respectively. Per-stem analysis reveals substantial variation in restoration difficulty across instruments, with bass averaging 4.59 dB across all teams, while percussion averages only 0.29 dB. The dataset, evaluation protocols, and baselines are available at https://msrchallenge.com/.

Nathan Paek, Yongyi Zang, Qihui Yang et al.

While sparse autoencoders (SAEs) successfully extract interpretable features from language models, applying them to audio generation faces unique challenges: audio's dense nature requires compression that obscures semantic meaning, and automatic feature characterization remains limited. We propose a framework for interpreting audio generative models by mapping their latent representations to human-interpretable acoustic concepts. We train SAEs on audio autoencoder latents, then learn linear mappings from SAE features to discretized acoustic properties (pitch, amplitude, and timbre). This enables both controllable manipulation and analysis of the AI music generation process, revealing how acoustic properties emerge during synthesis. We validate our approach on continuous (DiffRhythm-VAE) and discrete (EnCodec, WavTokenizer) audio latent spaces, and analyze DiffRhythm, a state-of-the-art text-to-music model, to demonstrate how pitch, timbre, and loudness evolve throughout generation. While our work is only done on audio modality, our framework can be extended to interpretable analysis of visual latent space generation models.

Ali Vosoughi, Yongyi Zang, Qihui Yang et al.

Room impulse response (RIR) generation remains a critical challenge for creating immersive virtual acoustic environments. Current methods suffer from two fundamental limitations: the scarcity of full-band RIR datasets and the inability of existing models to generate acoustically accurate responses from diverse input modalities. We present PromptReverb, a two-stage generative framework that addresses these challenges. Our approach combines a variational autoencoder that upsamples band-limited RIRs to full-band quality (48 kHz), and a conditional diffusion transformer model based on rectified flow matching that generates RIRs from descriptions in natural language. Empirical evaluation demonstrates that PromptReverb produces RIRs with superior perceptual quality and acoustic accuracy compared to existing methods, achieving 8.8% mean RT60 error compared to -37% for widely used baselines and yielding more realistic room-acoustic parameters. Our method enables practical applications in virtual reality, architectural acoustics, and audio production where flexible, high-quality RIR synthesis is essential.

Yongyi Zang, Zheqi Dai, Mark D. Plumbley et al.

We introduce Music Source Restoration (MSR), a novel task addressing the gap between idealized source separation and real-world music production. Current Music Source Separation (MSS) approaches assume mixtures are simple sums of sources, ignoring signal degradations employed during music production like equalization, compression, and reverb. MSR models mixtures as degraded sums of individually degraded sources, with the goal of recovering original, undegraded signals. Due to the lack of data for MSR, we present RawStems, a dataset annotation of 578 songs with unprocessed source signals organized into 8 primary and 17 secondary instrument groups, totaling 354.13 hours. To the best of our knowledge, RawStems is the first dataset that contains unprocessed music stems with hierarchical categories. We consider spectral filtering, dynamic range compression, harmonic distortion, reverb and lossy codec as possible degradations, and establish U-Former as a baseline method, demonstrating the feasibility of MSR on our dataset. We release the RawStems dataset annotations, degradation simulation pipeline, training code and pre-trained models to be publicly available.

Yongyi Zang, Jingyi Li, Qiuqiang Kong

Audio source separation aims to separate a mixture into target sources. Previous audio source separation systems usually conduct one-step inference, which does not fully explore the separation ability of models. In this work, we reveal that pretrained one-step audio source separation models can be leveraged for multi-step separation without additional training. We propose a simple yet effective inference method that iteratively applies separation by optimally blending the input mixture with the previous step's separation result. At each step, we determine the optimal blending ratio by maximizing a metric. We prove that our method always yield improvement over one-step inference, provide error bounds based on model smoothness and metric robustness, and provide theoretical analysis connecting our method to denoising along linear interpolation paths between noise and clean distributions, a property we link to denoising diffusion bridge models. Our approach effectively delivers improved separation performance as a "free lunch" from existing models. Our empirical results demonstrate that our multi-step separation approach consistently outperforms one-step inference across both speech enhancement and music source separation tasks, and can achieve scaling performance similar to training a larger model, using more data, or in some cases employing a multi-step training objective. These improvements appear not only on the optimization metric during multi-step inference, but also extend to nearly all non-optimized metrics (with one exception). We also discuss limitations of our approach and directions for future research.