每日 arXiv

音频 · 语音 · 音乐 · 声学

每日新爬取的 arXiv 论文，按评分排序。中文摘要帮你快速判断是否值得深读。

2026-07-12
日期2026-07-12
已评分—
均分—
最高—

Daily Papers — 2026-07-12

4 papers on audio, speech, music, and acoustics.

1. Data Augmentation for L2 English Speaking Assessment using TTS

Authors: Stefano Bannò, Penny Karanasou, Mengjie Qian, Kate M. Knill, Mark J. F. Gales

Categories: eess.AS Score: 5.74/10 (Obj:8 Id:5 Ind:6 Comp:6 Eff:5 Nov:5 Rep:7)

  • Strength: 系统性对比四种 speaker-text 匹配策略发现 proficiency 匹配最稳健，speechification 步骤使 ModernBERT PCC 从 0.517 提升至 0.744，且增强后两个评分器在 RealSpoken test set 上均获一致改善（wav2vec2 +0.053 PCC, ModernBERT +0.056 PCC）。
  • Weakness: 未与最直接的同类工作（Wang et al. [6], Voskoboinik et al. [8]）做定量对比，未尝试标准语音增强 baseline，仅单一小规模数据集（410 训练样本），核心 pipeline 与同团队前作 [10] 技术栈高度重叠。
  • Full review: Claude Code 全文七公理审稿

Automated assessment of second language (L2) speaking proficiency relies on large-scale annotated speech data, which remains scarce compared to widely available written learner corpora. A promising direction for addressing this imbalance is to use text-to-speech (TTS) and voice cloning to convert written L2 production into synthetic speech. However, written and spoken L2 differ fundamentally: spontaneous speech includes disfluencies and discourse markers, while writing is more planned and complex. This raises the question of what is required to generate synthetic L2 speech suitable for assessment. We address this through a systematic analysis of speaker-text relationships using COREFL, a publicly available corpus containing paired spoken and written responses from the same L2 learners to the same questions across modalities. In our proposed framework, we first address the structural differences between written and spoken language by transforming written responses into spoken-style transcripts (“speechification”) using a large language model. These transcripts are then converted into speech using a TTS/voice-cloning model. To assign a voice to each synthetic response, we investigate different speaker-text pairing strategies based on shared learner attributes (proficiency level, first language, both, or neither). We evaluate our data augmentation techniques on the language assessment task, with improvements shown in both wav2vec2 (audio-based) and ModernBERT (text-based) scoring systems. Results show that matching speakers and texts by proficiency level yields the most robust synthetic speech. Moreover, raw written text leads to a strong mismatch with spoken language, while speechification substantially reduces this gap and improves grading performance.

2. ECHOv2: Two-Level Band-Splitting Representation Learning for Anomalous Sound Detection

Authors: Yucong Zhang, Juan Liu, Ming Li

Categories: eess.AS | Submit to TASLP Score: 5.70/10 (Obj:8 Id:6 Ind:8 Comp:5 Eff:5 Nov:4 Rep:9)

  • Strength: 在同 backbone/同数据/同 compute 下通过 inter-band self-distillation 在 DCASE 2020–2025 六年数据集上实现 embedding-based overall AUC 62.63%（vs ECHO-Small 62.11%，p=0.0353 显著），代码和 benchmark 完全开源。
  • Weakness: 相对前作 ECHO-Small 的绝对提升仅 +0.52pp（embedding）/ +0.25pp（adaptation，不显著），inter-band 各组件均为已有 SSL 技术的组合，未排除额外 100k 步训练量的混淆。
  • Full review: Claude Code 全文七公理审稿

Machine anomalous sound detection (ASD) requires robust audio representations capable of capturing subtle deviations in machine sounds under limited supervision. Existing pre-trained audio backbones do not fully capture frequency-specific characteristics of machine sounds. To address this, we propose ECHOv2, a band-splitting model that learns localized intra-band representations to capture fine-grained spectral patterns while also incorporating a two-level self-distillation strategy with explicit inter-band supervision to model cross-frequency dependencies. The inter-band branch performs global context alignment and masked sub-band reconstruction, and multiple summary tokens are introduced for structured aggregation with controllable frequency granularity, enabling region-aware interaction across sub-bands during training. This design allows ECHOv2 to robustly handle diverse machine types and noisy operating conditions while maintaining stable representation quality. To enable fair and consistent evaluation of pre-trained audio backbones, we establish a unified ASD benchmark over DCASE 2020-2025 with two complementary protocols: embedding-based evaluation for frozen representation discriminability and adaptation-based evaluation for downstream transferability. Ablation studies confirm the effectiveness of intra-band learning, inter-band supervision, and structured aggregation granularity for robust ASD representation learning. These findings demonstrate that structured cross-band modeling provides a powerful and adaptable framework for ASD representation learning and can serve as a strong foundation for future research. The model and benchmark are fully open-sourced at https://github.com/yucongzh/ECHOv2 and https://github.com/yucongzh/ASD_Benchmark to promote reproducible research.

3. Dance to Music Generation leveraging Pre-training with Unpaired data and Contrastive Alignment

Authors: Ryota Kimura, Sangheon Park, Natalia Polouliakh, Taketo Akama

Categories: cs.SD, cs.MM, eess.AS | 7 pages, 1 figure Score: 5.50/10 (Obj:8 Id:5 Ind:6 Comp:4 Eff:5 Nov:4 Rep:9)

  • Strength: 在 AIST++ 上 BAS 从 0.194 提升至 0.234（达 ground-truth 87.3%），FAD 从 6.90 降至 4.56，代码完全开源，训练细节充分。
  • Weakness: 三组件拼装（MotionBERT + BeatDance-style contrastive + 标准 ControlNet）缺乏机制创新，sound quality MOS 2.65 劣于 baseline 2.88，消融显示移除对比预训练后多数客观指标反而更好，未与 MoMu-Diffusion (NeurIPS 2024) 直接比较。
  • Full review: Claude Code 全文七公理审稿

Dance-to-music generation is a promising task for applications such as choreography support and automatic accompaniment, where temporal coordination between body movement and sound is essential. In particular, using human joint positions as the motion representation is attractive because they explicitly capture body dynamics while being lightweight, privacy-preserving, and easy to integrate with motion capture and pose-estimation pipelines. A central challenge in this setting, however, is the scarcity of high-quality paired dance-music data, since collecting accurately synchronized pairs is costly and often constrained by copyright and performance rights. This makes it difficult to train end-to-end models solely from paired data. To address this issue, we propose a dance-conditioned music generation framework that efficiently exploits both unpaired and paired data. Our method combines pretrained unimodal encoders for motion and music, beat-guided contrastive pretraining to align their feature spaces, and a ControlNet-style conditioning module on top of a pretrained text-to-audio diffusion model. Experiments on AIST++ demonstrate that the proposed techniques improve both dance-music alignment and audio quality, as confirmed by quantitative and qualitative evaluations. Compared to a state-of-the-art method, our approach achieves superior dance alignment performance and competitive audio quality. Code is available at https://github.com/kmraven/AudioLDM-ControlNet .

4. An Objective Intelligibility Metric Evaluation on Spanish Speech

Authors: Iván López-Espejo, Jesper Jensen

Categories: eess.AS | Submitted to IberSPEECH 2026 Score: 4.84/10 (Obj:5 Id:4 Ind:8 Comp:5 Eff:4 Nov:4 Rep:8)

  • Strength: SpInt 是首个公开的西班牙语语音可懂度数据集（5,148 条语音，26 名母语者，Zenodo 公开），人类评分采集通过独立听力实验完成，所有被评估 metric 均为第三方独立开发，可复现性高。
  • Weakness: 核心发现被作者自述为 “unsurprising” 且已被现有文献覆盖；数据集仅覆盖 1 种噪声/1 个说话人/3 种处理条件，泛化性受限；未评估最新 NR-OIM SOTA (TSIP-Net)；对 reference-based 优于 no-reference 的因果归因缺乏控制实验。
  • Full review: Claude Code 全文七公理审稿

Objective intelligibility metrics (OIMs) enable fast and low-cost evaluation of speech intelligibility and are widely used in speech technology assessment. In this study, we evaluate five reference-based OIMs (STOI, ESTOI, STGI, HASPI, and SIIB) and two deep learning-based no-reference metrics (MOSA-Net+ and W2V-SIP) on SpInt, a new Spanish speech intelligibility dataset. Our results show that reference-based OIMs consistently outperform modern data-driven no-reference approaches, which degrade notably under training-test acoustic mismatches such as language mismatch. This effect is particularly relevant in our scenario, as none of the evaluated metrics were exposed to Spanish speech data during development. Consequently, to foster research on more robust and generalizable no-reference OIMs, SpInt is released publicly.