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。

2026-07-07
日期2026-07-07
已评分—
均分—
最高—

Daily Papers — 2026-07-07

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

1. Compress the Cache, Not the Speech Embedding: KV Compression for Efficient Speech LLMs

Authors: Ke-Han Lu, Keqi Deng, Ruchao Fan, Rui Zhao, Jinyu Li

Categories: eess.AS, cs.SD | Submitted to SLT2026 Score: 7.37/10 (Obj:8 Id:8 Ind:8 Comp:8 Eff:8 Nov:6 Rep:6)

  • Strength: 中高 — 方法简洁有效，问题重构（”compress the cache, not the embedding”）有启发性，R=4 时压缩无性能损失甚至略有提升是一个有价值的实用结论。但核心机制（learned pooling for KV compression）在文本 LLM 中已有充分探索。
  • Weakness: KV cache 压缩在文本 LLM 中已有大量工作：DynamicKV (EMNLP 2025, task-aware adaptive compression)、Nawrot et al. 2024 (dynamic KV)、Chevalier et al. 2023 (pooling KV tokens)、Wang et al. 2024 (model compression)。
  • Full review: Claude Code 全文七公理审稿

Speech large language models (Speech LLMs) typically encode speech into sequences far longer than text, creating a major efficiency bottleneck during autoregressive decoding. A common remedy is to compress the speech sequence at the adapter level to remove temporal redundancy before it enters the LLM; however, such early downsampling risks discarding fine-grained information that cannot be recovered. We propose SpeechKV, which applies a learned pooling to the KV cache of speech tokens inside the LLM. This design allows the LLM to fuse speech and text internally while directly accelerating decoding. Trained on 71K hours of speech data, SpeechKV compresses the speech to approximately text-level granularity yet maintains performance on par with or even slightly better than the uncompressed baseline, with relative gains of 6.6% on out-of-domain entity recognition and 2.3% on OpenASR, while delivering at least 1.49 times decoding speedup that scales with audio length.

2. LEGATO 2: Toward Multimodal Sheet Music Recognition and Understanding

Authors: Guang Yang, Brian Siyuan Zheng, Victoria Ebert, Noah A. Smith

Categories: cs.CV, cs.AI | 23 pages. Equal contribution: Guang Yang and Brian Siyuan Zheng Score: 7.20/10 (Obj:9 Id:6 Ind:7 Comp:6 Eff:9 Nov:6 Rep:7)

  • Strength: Legato 2 在所有六个 OMR 数据集上均优于先前 SOTA (Legato 1)（OMR-NED 提升高达 46%），通过逐系统自回归处理实现自然多页扩展，并引入了首个神经 OMR 文本识别（总 CER 24.8 对比 Audiveris 73.8），在 SSMR-Bench 上达到 92.7% 的准确率。
  • Weakness: 消融实验未将系统分割的两个机制（有效分辨率提升 vs. 自回归上下文条件）分离开来，也未在相同系统级分辨率下与无上下文基线进行比较，这使得性能提升的实际原因尚不明确。
  • Full review: Claude Code 全文七公理审稿

We propose a novel pipeline, Legato 2, for extracting symbolic notation and semantic knowledge from images of sheet music. Legato 2 features the first large-scale neural model for optical music recognition (OMR) to operate sequentially on a system-by-system basis, following the horizontal lines of notation as they are read on the page, rather than treating the page as an undifferentiated image, enabling better scaling to arbitrarily long inputs. It is also the first OMR model capable of generating symbolic transcriptions that include embedded textual content, such as titles and annotations. The pipeline combines system-level segmentation with an autoregressive vision-LM to capture both local notation details and score structure. Across multiple datasets, Legato 2 consistently outperforms prior state of the art. We also show that symbolic transcriptions complement visual inputs for frontier language models, improving their interpretation of dense musical documents. Legato 2 establishes new state-of-the-art performance in both OMR and downstream sheet music understanding.

3. Escaping the Procrustean Bed: Groupwise Orthogonal Connectors for Audio-Language Models

Authors: Ho-Lam Chung, Ke-Han Lu, Yi-Cheng Lin, Guan-Ting Lin, Yiming Chen et al.

Categories: cs.SD Score: 7.10/10 (Obj:9 Id:8 Ind:6 Comp:8 Eff:7 Nov:6 Rep:6)

  • Strength: ORCA 用极简的分组正交约束（参数量和推理成本与 baseline 完全匹配），在 controlled comparison 下将 SAKURA multi-hop 从 48.8% 提升到 75.2%，超越 8B Audio Flamingo-3 的 49.0%，且 cross-speaker variance 提升 75× 作为纯几何约束的 downstream effect。
  • Weakness: 方法仅在 paralinguistic reasoning 上有显著效用，MMAU general audio understanding 上 Speech 指标反降 2.4 points；未提供代码/checkpoint，且 SAKURA benchmark 作者与方法作者重叠存在独立性隐患。
  • Full review: Claude Code 全文七公理审稿

Audio-language models compress a speech encoder’s output through a Querying Transformer (Q-Former) connector before feeding it to a large language model. We identify two failures in this compression. The connector’s output vectors collapse to a single direction, and different speakers produce nearly indistinguishable outputs, with paralinguistic cues such as speaker identity, gender, and prosody lost along the way. Our method, ORCA, reverses this collapse by splitting the queries into groups whose outputs are constrained to point in different directions. On SAKURA multi-hop reasoning, ORCA gains 26.4 points over an identically trained 4B baseline, reaching 75.2% (vs. 49.0% for the 8B Audio Flamingo-3). At the connector level, the same change cuts query redundancy by 12x and raises cross-speaker variance by 75x.

4. ForestIR: Physics-Informed Forest Sound Simulation for Array-Based Bioacoustic Remote Sensing

Authors: Xin Shen, Jennifer N. Kampe, Changwoo J. Lee, Braden Scherting, Panu Somervuo et al.

Categories: eess.AS, cs.SD Score: 7.10/10 (Obj:9 Id:6 Ind:9 Comp:6 Eff:8 Nov:6 Rep:9)

  • Strength: 中 — 将多个成熟声学建模组件整合为参数可控的端到端系统，对定位研究社区有实用价值。但缺乏逐组件 ablation 和对简化假设（如地面反射幅度乘数）的独立验证。
  • Weakness: 论文与 legacy model (Kaneko & Gamper 2021) 进行了系统比较，这是正确的最简 baseline 选择。在定位敏感性实验中，论文通过控制变量法（仅改变树木布局/仅改变温度）隔离了关键变量，实验设计合理。
  • Full review: Claude Code 全文七公理审稿

Microphone array-based passive acoustic monitoring is increasingly used for biodiversity sensing in forests. However, design and evaluation of array systems and configurations remains difficult since field recordings are costly, difficult to reproduce, and provide limited control over forest and atmospheric conditions. We present ForestIR, a physics-informed and reproducible simulation framework that links forest and environmental conditions to microphone-array recordings for bioacoustic remote sensing. Through a more realistic sound propagation method and a systematic control over array design and environmental factors, ForestIR provides a practical simulation framework for optimizing array-based monitoring systems, especially for sound source localization purposes. ForestIR generates source-microphone impulse responses (IRs) under user-controlled forest and atmospheric conditions, and renders synthetic array recordings by convolving test signals with controlled background noise. We evaluate and demonstrate realistic features of ForestIR through experiments based on localization sensitivity to forest layout and atmospheric conditions, and also comparison between simulated IRs with sine-sweep IR measurements from a field experiment. ForestIR provides a practical way to test how forest and ground conditions, atmospheric state, and array geometry affect bioacoustic localization, and can support microphone-array design, robustness testing, and synthetic-data generation for passive acoustic monitoring.

5. Hierarchical Acoustic-Semantic Modeling: Modality Separation and Semantic Coherence for Full-Duplex SLMs

Authors: Zhenyu Liu, Xuanyu Zhang, Yunxin Li, Qixun Teng, Shenyuan Jiang et al.

Categories: cs.CL | 22 pages, 9 figures, Fixed author Score: 6.95/10 (Obj:8 Id:8 Ind:5 Comp:6 Eff:8 Nov:6 Rep:8)

  • Strength: 中高 — 两个组件都有 ablation 支持必要性，效果提升显著且全面。但组件本身（shared backbone + task heads, dense text supervision）是已知策略的组合，压缩价值有限。
  • Weakness: 存在两个独立性隐患
  • Full review: Claude Code 全文七公理审稿

Developing seamless, high-performance, native intelligent full-duplex Spoken Language Models (SLMs) remains a critical challenge and long-standing goal for the speech and NLP community. Despite notable progress, recent endeavors are fundamentally constrained by severe modality interference, which causes substantial knowledge degradation and compromises semantic integrity – ultimately making full-duplex SLMs feel unnatural and unintelligent. In this paper, through an exhaustive fine-grained analysis of model optimization dynamics, we uncover the root cause of such performance degradation, revealing that modality interference arises from inherent gradient conflicts between acoustic and semantic modeling when the two modalities are forced to share a deep parameter space. Guided by this key insight, we introduce Lychee-FD, a native end-to-end full-duplex framework designed to mitigate modality interference. Importantly, we propose a hierarchical parameter separation strategy that decouples conflicting modalities in deep layers while preserving cross-modality coherence via a dedicated semantic alignment channel. Extensive experiments on multiple full-duplex benchmarks demonstrate that our method significantly advances the state of the art, yielding substantial improvements in both speech intelligence (+7.4% on Spoken QA) and full-duplex interaction fluidity (+28.5% on FullDuplexBench 1.5) without compromising inference efficiency. To the best of our knowledge, this work is the first to achieve two key advances: 1) uncovering and elucidating the root cause of modality interference in full-duplex SLMs, and 2) designing an elegant hierarchical model together with a practical solution for seamless, high-performance, native intelligent full-duplex SLMs.

6. WordVoice: Explicit and Decoupled Multi-Dimensional Word-Level Control for LLM-Based TTS

Authors: Sihang Nie, Jinxin Ji, Xiaofen Xing, Deyi Tuo, Chengbin Jin et al.

Categories: eess.AS, cs.CL, cs.SD | 10 pages, 4 figures, 6 tables; Preprint Score: 6.95/10 (Obj:8 Id:5 Ind:5 Comp:6 Eff:8 Nov:8 Rep:8)

  • Strength: 中高 — bound-token + WV-FM的双阶段设计有效解决了token-to-waveform的resolution mismatch，解耦实验和ablation证明了模块分工的合理性，free/control双模式推理具有实用价值。
  • Weakness: 论文的ablation存在缺口。（1）baseline选择：主实验仅与CosyVoice3（backbone）和MagicTTS（单维度时间控制）比较。WeSCon被排除的理由是”abstract emotion control lacks deterministic mappings to specific acoustic attributes”——这一排除逻辑合理但意味着缺少一个word-level控制的直接竞争者。
  • Full review: Claude Code 全文七公理审稿

While recent Large Language Model (LLM)-based Text-to-Speech (TTS) systems have achieved remarkable naturalness, they predominantly rely on implicit end-to-end generation paradigms, resulting in coarse-grained control. In scenarios demanding precise stylistic interventions and strict temporal alignment, such as audiobook narration and video dubbing, the inability to explicitly manipulate word-level acoustic attributes remains a critical bottleneck. This limitation is primarily amplified by the severe scarcity of fine-grained annotated datasets and the architectural challenge of integrating multi-dimensional control signals into discrete autoregressive generation. To address this, we propose a unified framework for highly precise word-level control. First, we construct WordVoice-5A, a massive 4.7k-hour bilingual dataset featuring five-dimensional word-level annotations (duration, boundary, energy, pitch and tone) developed through a rigorous linguistically-guided pipeline. Second, we introduce WordVoice to transform the implicit generation process into an explicit, highly controllable paradigm. Specifically, we introduce a bound-token mechanism within the LLM to formulate an explicit ``acoustic planning’’ process, enabling adaptive multi-task prosodic planning and flexible manual intervention. Furthermore, we augment the token-to-waveform stage with a fine-grained acoustic modulation module, bridging the resolution gap to strictly align word-level attributes between highly compressed discrete tokens and continuous waveforms. Extensive experiments demonstrate that WordVoice achieves superior, decoupled control over multiple acoustic dimensions while maintaining competitive zero-shot synthesis stability. The code and audio samples are publicly available at https://xxh333.github.io/wordvoice-demo/.

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

Authors: Ho-Lam Chung, Kuan-Po Huang, Bo-Ru Lu, Hung-yi Lee

Categories: cs.SD Score: 6.70/10 (Obj:9 Id:6 Ind:5 Comp:8 Eff:8 Nov:6 Rep:5)

  • Strength: 中 — SR-FD 设计轻量、推理零开销、三 target 角色明确，是一个实用的 TTS fine-tuning 工具。但未证明在 VoxCPM2 之外的基座上泛化。
  • Weakness: 存在两处独立性隐患：(1) Whisper 同时用于 SR-FD 训练目标和 WER 评测——SR-FD 的 low-step Whisper anchor target 使用 Whisper encoder 提取特征，而 Seed-TTS 的 WER 评测也依赖 ASR 系统（虽然上游 scorer 的具体 ASR 模型未完全披露，但 Whisper 是最可能的选择）。
  • Full review: Claude Code 全文七公理审稿

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.

8. Music I Care About: Automated Multimodal Benchmarking of LLM Music Perception Skills on (Almost) Any Music

Authors: Tomáš Sourada, Katia Vendrame, Jan Hajič

Categories: cs.SD | 9 pages, 4 figures, 4 tables Score: 6.50/10 (Obj:8 Id:6 Ind:8 Comp:5 Eff:6 Nov:6 Rep:8)

  • Strength: 程序化 pipeline 设计简洁合理，no-input/noise-input 双 ablation 是有效的验证策略，但 distractor 质量控制仍有缺口，数据集覆盖范围过窄。
  • Weakness: 框架的核心设计是：问题模板 + 本体（ontology）+ MusicXML → 程序化 ground truth 提取 → 跨模态复制。这一设计是简洁的，没有不必要的复杂模块。
  • Full review: Claude Code 全文七公理审稿

Music represents a cornerstone of human culture, existing digitally across diverse modalities, including audio, symbolic encodings (e.g., MIDI, MusicXML), and sheet music. Despite the advancement of Multimodal Large Language Models (MLLMs), current music benchmarks face three major limitations. First, large static benchmarks are resource-intensive to evaluate, and it remains unclear how their results transfer to diverse kinds of music beyond those included in the benchmark. Second, benchmarks claiming to measure “music understanding” often fail to require music perception. Third, they do not support systematic performance comparisons across musical modalities. To overcome these issues, we introduce the Music I Care About Meta-Benchmark (MusICA-MetaBench), a framework that automatically derives on-demand benchmarks directly from user-provided data. By leveraging structured symbolic representations (e.g., MusicXML) and our pre-defined question templates, we build multiple-choice question-answer pairs that probe music perception competencies, aligned with music pedagogy, across audio, music notation images, and symbolic files. We demonstrate our framework with the ChoraleBricks dataset, and experimentally determine benchmark sizes that ensure statistically reliable model comparisons for this setup. By comparing against text-only and white-noise baselines, we show our questions do measure music perception. Ultimately, MusICA-MetaBench represents a significant advancement in the cross-modal assessment of music perception for MLLMs. By proposing a dataset-specific benchmarking paradigm, it enables efficient on-demand evaluation of music perception capabilities.

9. Precise Video-to-Audio Generation with Cross-Modal Alignment in Latent Space

Authors: Thanh V. T. Tran, Ngoc-Son Nguyen, Luong Tran, Long-Khanh Pham, Paarth Neekhara et al.

Categories: cs.MM, cs.SD | Accepted to ECCV 2026 Score: 6.40/10 (Obj:8 Id:6 Ind:5 Comp:8 Eff:8 Nov:6 Rep:6)

  • Strength: Flowley 通过 PSCA 机制在零额外计算开销下实现音视频时间对齐，169M 参数在 VGGSound 上 5/7 指标 SOTA，SoundCap 使 MMAudio KAD 改善 31.6%，零样本 IS 超过 13B 的 Movie Gen Audio。
  • Weakness: PSCA 之外的核心架构为 MM-DiT+DiT 直接组合且 Flowley 在 Align Acc 上不及使用同评测编码器的 Frieren，SoundCap 缺乏独立人类标注校验且训练技巧未充分消融。
  • Full review: Claude Code 全文七公理审稿

Video-to-audio (V2A) generation aims to synthesize realistic audio that is both semantically consistent with and temporally synchronized to a silent video. Despite recent progress, many methods still rely on multi-stage training, resulting in high computational costs and long runtimes, or transform visual input into text to leverage pretrained text-to-audio models, sacrificing fine-grained temporal cues. To overcome these limitations, we propose Flowley, an end-to-end, single-stage training architecture that produces soundtracks by combining visual features with textual prompts. Crucially, we introduce Progressive Soft-masked Cross-Attention, which embeds audio-visual synchronization directly within its attention mechanism, adding zero additional computational cost compared to standard attention layers. We further observe that existing V2A benchmarks lack sound-oriented descriptive captions, which can potentially degrade the quality of the synthesized audio. To remedy this, we propose SoundCap, a plug-and-play pipeline for creating detailed, sound-aware captions that guide the model. Remarkably, without integrating any pretrained audio-visual alignment modules, Flowley achieves state-of-the-art performance on VGGSound across multiple metrics. Moreover, by incorporating SoundCap, we further exceed the performance of the strongest existing close-sourced methods in terms of audio quality in the zero-shot setting.

10. Goodbye Equal Error Rate, Hello Local Information Disclosure: Evaluating Voice Anonymisation against 1-to-N Linkage Threats

Authors: Dāvis Šterns, Konstantinos Drossos, Natasha Fernandes, Tom Bäckström, Catuscia Palamidessi

Categories: eess.AS Score: 6.20/10 (Obj:8 Id:6 Ind:8 Comp:6 Eff:8 Nov:5 Rep:5)

  • Strength: 基于 PMI 的 LID 指标在信息论上严谨，清晰证明 VPC 2024 顶级系统在 EER≈0.48 时仍有 >63% 试验泄露正信息、LIDmax > 1 bit，暴露了 EER 在 1-to-N 威胁模型下的系统性评价缺陷。
  • Weakness: 核心方法是在 SRD（共同作者 Bäckström 的前序工作）基础上的连续化扩展，缺少校准管道各组件的消融实验，且无代码开源；N=29 的 enrolment 规模偏小，未讨论大 N 行为。
  • Full review: Claude Code 全文七公理审稿

Voice anonymisation aims to protect speaker identity. Currently, its empirical privacy evaluation heavily relies on the Equal Error Rate (EER). Originally designed for biometric verification, EER aggregates scores globally, implicitly assuming an attacker is only trying to verify if two specific voice samples match (a 1-to-1 comparison). This introduces a threat model mismatch with real-world database linkage attacks, where an attacker searches across a fixed set of N enrolled identities (a 1-to-N closed-set search), allowing global averages to obscure localised privacy failures. While recent 1-to-N metrics address this aggregation issue, they abstract away the magnitude of the biometric evidence. In this paper, we propose a modular, information-theoretic evaluation framework explicitly designed for the 1-to-N linkage threat model. Within this framework, our core metric, Local Information Disclosure (LID), quantifies the exact privacy loss of a single trial utterance in bits by calibrating its raw similarity scores into the attacker’s posterior confidence for each enrolled identity. Evaluating top-performing systems from the VoicePrivacy 2024 Challenge reveals that systems exhibiting near-perfect EERs (48 %) can still suffer from localised vulnerabilities with worst-case disclosures reaching 1 bit per trial utterance (effectively doubling the attacker’s success rate over a random guess). We demonstrate that adopting localised privacy metrics is essential for capturing worst-case risks and aligning with strict privacy regulations.

11. Gradient-Based Speech-to-Text Alignment for Any ASR Model: From CTC to Speech LLMs

Authors: Albert Zeyer, Ralf Schlüter, Hermann Ney

Categories: cs.CL, cs.AI, cs.CV, cs.LG Score: 6.00/10 (Obj:8 Id:6 Ind:8 Comp:5 Eff:6 Nov:5 Rep:8)

  • Strength: 单一训练自由的输入梯度显著性方法在 16 个模型（4 个家族，含 Speech LLM）上均产生可用词对齐，在流式模型和 Canary-Qwen 上超越原生对齐（FastConformer streaming: 梯度 127ms vs 后验 357ms WBE on TIMIT），并在 Whisper-large-v3 最优编码器深度上超越 cross-attention DTW（33ms vs 42ms on TIMIT）。
  • Weakness: 方法是经典 saliency map (Simonyan 2014) + NMT 梯度词对齐 (Ding 2019) + CTC 拓扑 DP 解码的组合，引入大量超参数且未充分解释机制；大多数场景下绝对效果远落后于 GM-HMM/MFA 金标准，作者自己也承认不提议作为实用对齐器；遗漏 WhisperX 强制对齐管线作为直接 baseline。
  • Full review: Claude Code 全文七公理审稿

Speech-to-text alignment means finding the temporal boundaries of each word in the audio. Some models provide such an alignment directly and others do not. Connectionist temporal classification (CTC) and transducer models have an alignment by construction, whereas attention-based encoder-decoders (AED) and speech large language models (LLMs) do not, and their word timings are usually read off the attention weights instead. All of these signals live on the encoder frame grid, which bounds their temporal precision. We study a generic gradient-based alignment that applies to any differentiable ASR model. We take the gradient of each teacher-forced token log probability with respect to the input, reduce it to a per-frame saliency, and decode the resulting matrix into word boundaries with a single dynamic-programming pass. The method needs no training, no model modification and no alignment heads, works across all model families including the speech LLMs, and aligns on the input grid rather than on the coarser encoder grid. We evaluate it on sixteen models from four families, on read (TIMIT) and spontaneous (Buckeye) speech, each against the model’s own native or attention-based alignment. We find that the gradient yields a usable alignment for every model, that it is usually somewhat behind a strong native aligner but better where the native alignment is weak, as for the streaming models, and that its main disadvantage is the cost of one backward pass per token.

12. InsideSSL: Understanding Self-Supervised Speech Representations using a Model-Centric Perspective

Authors: Samir Sadok, Xavier Alameda-Pineda

Categories: cs.SD | 10 pages - 9 figures - accepted at INTERSPEECH 2026 (long paper track) Score: 5.90/10 (Obj:8 Id:5 Ind:8 Comp:5 Eff:5 Nov:5 Rep:8)

  • Strength: GCM 作为首个 cross-layer 生成兼容性分析工具揭示了语音 SSL 中 stable phonetic core 和 hierarchical pruning 的结构，且 5 个模型间的一致性高（curvature 相关 >0.96），实验全可复现。
  • Weakness: per-layer 三指标直接迁移自 Skean et al. (2025) 的 NLP 工作且实现复用其仓库，主要下游任务发现大量确认 Pasad et al. 已有结论，全部实验仅限 LibriSpeech 单一数据集，缺乏因果隔离实验。
  • Full review: Claude Code 全文七公理审稿

Self-supervised learning (SSL) models, such as Wav2Vec2, HuBERT, and WavLM, have become foundational across a wide range of speech and audio tasks. Despite their success, understanding their internal layer-wise dynamics remains an ongoing challenge. To address this, we propose a two-part model-centric framework called InsideSSL. First, we establish a task-agnostic analysis from three intrinsic per-layer perspectives: compression (entropy), geometry (curvature), and robustness to perturbations. We show that varying training objectives induce distinct regimes of acoustic compression and manifold unfolding. Second, we introduce the cross-layer Generative Compatibility Matrix (GCM) to evaluate functional transferability, exposing stable phonetic cores, identity volatility, and deep-layer semantic pruning. In addition to these evaluations, linear probing connects the model-centric perspective to downstream tasks, demonstrating how layer topology dictates phoneme, pitch, and speaker encoding.

13. Determinantal point process sampling for bioacoustic active learning

Authors: Hugo Magaldi, Gabriel Dubus

Categories: cs.SD, cs.LG | BioDCASE Challenge 2026 - Task 4 Active learning. Ranked 2/14 Score: 5.90/10 (Obj:8 Id:6 Ind:9 Comp:5 Eff:6 Nov:5 Rep:9)

  • Strength: 在受控 BioDCASE 2026 框架下，CARE-DPP 通过 DPP 批次多样化 + 自适应批次调度，在四个生物声学数据集上实现 mean AULC 0.50（vs CoreSet 0.46，+8.7%），消融实验清晰识别 DPP 为最大贡献组件（移除后 AULC 降 0.038）。
  • Weakness: 方法为四个已有组件的组合（DPP batch AL 来自 Biyik 2019、类别平衡来自 Cui 2019、退火调度和自适应批次为标准启发式），缺少与原始 DPP-AL baseline 的直接对比，且 baseline 使用固定批次而本文用自适应批次导致比较不完全公平，同期同团队论文（arXiv 2607.04868）的贡献重叠度未澄清。
  • Full review: Claude Code 全文七公理审稿

Eco-acoustic monitoring generates vast volumes of audio data, making active learning a promising approach for reducing annotation effort while efficiently training reliable biodiversity classifiers. This report presents CARE-DPP, a batch active-learning acquisition method submitted to BioDCASE Active Learning for Bioacoustics 2026 challenge. The method combines class-balanced predictive uncertainty with embedding-space novelty, while a determinantal point process (DPP) objective selects a high-quality and non-redundant acquisition batch. The uncertainty-novelty balance is annealed over the annotation budget: early cycles emphasize geometric coverage, whereas later cycles increasingly exploit classifier uncertainty. To mitigate unreliable early scores, the DPP candidate pool mixes top-quality candidates with a decreasing proportion of random exploration. An adaptive acquisition schedule uses smaller batches early and larger batches later. Evaluated over five repeats on the BirdSet HSN, POW and UHH subsets and on ATBFL, CARE-DPP obtains a mean development AULC of 0.50 for macro mAP, compared with 0.46 for the official CoreSet baseline. Ablations identify DPP batch diversification and the adaptive acquisition schedule as the largest contributors.

14. Flow Matching-Based Speech Source Separation with Best-of-N Biometric Sampling

Authors: Anastasia Zorkina, Alexandr Anikin, Nikita Khmelev, Anastasiya Korenevskaya, Sergey Novoselov et al.

Categories: cs.SD | Accepted at the ICML 2026 Workshop on Machine Learning for Audio Score: 5.70/10 (Obj:8 Id:5 Ind:4 Comp:5 Eff:6 Nov:6 Rep:5)

  • Strength: 中 — 生物特征排序+选择+对齐的统一思路有实用性，但每个组件的单独创新增量有限，整体是工程集成而非机制创新。
  • Weakness: 存在明显的循环依赖嫌疑
  • Full review: Claude Code 全文七公理审稿

Single-channel speech separation remains challenging for real-world deployment due to source permutation ambiguity, sampling variability of generative models, and the difficulty of processing long recordings with chunk-wise inference. We address these issues with a conditional flow-matching-based method that produces an ordered two-source output conditioned on the mixture. A frozen speaker encoder defines the source order during training and is reused at inference for biometric best-of-$N$ candidate selection and chunk-level channel alignment. We evaluate separation quality on Libri2Mix benchmark using SI-SDR, PESQ, and ESTOI, and measure downstream impact using cpWER for automatic speech recognition and EER for speaker verification. The results show that the proposed Transformer U-Net variant is competitive with strong baselines in objective separation metrics and achieves the lowest downstream automatic speech recognition and speaker verification error rates in all evaluated settings.

15. BlueMagpie-TTS: A Token-Efficient Tokenizer, Language Model, and TTS for Taiwanese-Accent Code-Switching Speech

Authors: Ho Lam Chung, Bo-Xuan Zheng, Cheng-Chieh Huang, Cheng-Han Chang, Jung-Ching Chen et al.

Categories: cs.SD, cs.CL Score: 5.70/10 (Obj:8 Id:6 Ind:5 Comp:5 Eff:7 Nov:4 Rep:8)

  • Strength: 通过 controlled comparison（同 acoustic stack、同 voice fine-tuning，仅前端不同）证明台湾本地化 LM 前端将 CER 从 6.43% 降至 4.81%（相对降低 25.2%），且 65.6% 盲听偏好胜出；PangolinTokenizer 在 8 个 tokenizer 中以最小词表（114,822）达到最低 token rate（0.485 tok/char）。
  • Weakness: 系统是标准组件（byte-level BPE + hybrid LM + reused acoustic stack + linear bridge）的工程拼装，无新机制；未与同实验室的 BreezyVoice (arXiv:2501.17790) 对比，tokenizer 与 LM 贡献未分离，测试集为内部非公开。
  • Full review: Claude Code 全文七公理审稿

Off-the-shelf TTS systems are poorly adapted to Taiwanese Mandarin. Their accent defaults to other Mandarin variants, their tokenizers over-segment common Taiwanese text, and their pronunciation degrades at code-switching boundaries where Chinese and English alternate within one utterance. These problems share one root: the text side lacks adaptation to the Taiwanese context. We address the text side from the bottom up. PangolinTokenizer, a byte-level BPE tokenizer trained on Taiwan-context data, reaches the lowest token rate (0.485 tokens/character) with the smallest vocabulary among nine tokenizers. Barbet, a billion-parameter Traditional-Chinese language model trained on PangolinTokenizer, serves as the text-semantic frontend and ranks first among comparable public models on a 14-task evaluation. BlueMagpie-TTS attaches Barbet to the pretrained acoustic stack of VoxCPM2 through a learned bridge, keeping the acoustic stack fixed. On a 1000-sentence Taiwan-localized test set, it lowers CER from 11.45% to 4.81% and WER from 14.83% to 5.36%, relative reductions of 58.0% and 63.9%. In a blind listening study on 500 of these sentences with ten listeners, 65.6% of majority votes prefer BlueMagpie-TTS.

16. From Sinhala to Dhivehi: Cross-Lingual Transfer Learning for Low-Resource Speech Recognition

Authors: Lukmal Ilyas, Nevidu Jayatilleke

Categories: cs.CL | 7 pages, 1 figure, 8 tables, Accepted paper at the 12th International Moratuwa Engineering Research Conference (MERCon) 2026 Score: 5.53/10 (Obj:6 Id:5 Ind:8 Comp:5 Eff:5 Nov:4 Rep:8)

  • Strength: 系统性比较了 5 种迁移策略（17 项实验），用 Turkish 控制实验有效隔离了语言相关性效应，最佳系统达到 12.89% WER，且代码完全开源。
  • Weakness: 迁移策略本身仅贡献 0.61 个绝对 WER 点改善（vs KenLM 的 27.77 点），CPT 实验存在架构不一致（XLS-R vs Wav2Vec2-BERT），未与 MMS 1B 和 Whisper 等关键 baseline 比较。
  • Full review: Claude Code 全文七公理审稿

Dhivehi, the national language of the Maldives, is currently under-resourced for automatic speech recognition (ASR) and other NLP tasks. This study investigates whether cross-lingual transfer learning from Sinhala, a linguistically related, relatively well-resourced Insular Indo-Aryan language, can improve Dhivehi ASR. We conduct seventeen experiments across five transfer learning paradigms: Dhivehi-only baselines, sequential fine-tuning, multilingual fine-tuning, continual pre-training, and a control using Turkish as an unrelated language. The strongest system, continual pre-training on Sinhala followed by fine-tuning on Dhivehi with KenLM, achieves 12.89% WER and 2.70% CER, outperforming the Dhivehi-only baseline by 13.50% WER and 3.02% CER. However, the adaptation strategy and decoding configuration are equally critical for a successful transfer learning experiment. We conduct seventeen controlled experiments spanning five transfer learning paradigms: Dhivehi-only baselines, sequential fine-tuning, multilingual fine-tuning, continual pre-training, and a control experiment using Turkish as an unrelated language. The strongest system, continual pre-training on Sinhala followed by fine-tuning on Dhivehi with KenLM, achieves 12.89% WER and 2.70% CER, outperforming the Dhivehi-only baseline by 13.50% WER and 3.02% CER. The Turkish control experiment confirms that observed improvements stem from linguistic relatedness; adaptation strategy and decoding configuration are also critical.

17. Andha-Dhun: A First Look at Audio Descriptions in Hindi

Authors: Ritabrata Chakraborty, Divy Kala, Nisheeth Bhooshan Gupta, Ganji Sreeram, Pailla Balakrishna Reddy et al.

Categories: cs.CV | Accepted to NCVPRIPG 2026, Download data at https://github.com/katha-ai/AndhaDhun-HindiAD Score: 5.53/10 (Obj:8 Id:5 Ind:4 Comp:5 Eff:5 Nov:6 Rep:6)

  • Strength: 低 — 全部使用已有方法组件，无新方法、新指标或新框架；两种 pipeline 是已有方法的直接组合。
  • Weakness: 存在明显的循环依赖问题。Gemini-3.1-Pro 同时用于：(1) 数据集转录中的 dialog/AD 分类；(2) Dense-to-Hindi 方法中的 AD 生成；(3) CSI 识别；(4) CSI 对应分类（same/partial/different）；(5) LLM-AD-Eval 的 judge。当 Gemini 既是生成者又是评判者时，评分的独立性严重受损。
  • Full review: Claude Code 全文七公理审稿

Audio Descriptions (ADs) narrate visual content for Blind and Low Vision (BLV) audiences during gaps in audiovisual media. There is growing momentum around ADs in movies and TV shows, and with mandates from India’s Central Board of Film Certification (CBFC), there is a need to expand ADs beyond English. Yet, there is no work that generates ADs for any Indian language. To address this gap, we present the first systematic study of ADs in Hindi, contributing to aspects such as data, generation, and evaluation. We introduce Andha-Dhun, the first dataset of human-authored Hindi ADs collected from 8 full-length movies. We explore two approaches for generating ADs in Hindi: (i) directly from English dense video descriptions, and (ii) translating English ADs into Hindi. We evaluate these approaches using perplexity and LLM-as-a-judge metrics to assess fluency and quality respectively. We also analyze movies that have both English and Hindi human-authored ADs and find that naive translation introduces artifacts and narrows diversity compared to original Hindi ADs. Direct machine translation fails to adapt cultural references, while human-translated ADs do better but still fall short. Our findings emphasize that the purpose of Hindi ADs is accessibility for Indian BLV audiences, and that this requires adapting content for the audience more than strict fidelity to the source.

18. Few-Shot Class-Incremental Audio Classification Using Pseudo-Incrementally Trained Embedding Learner and Continually Updated Stochastic Classifier

Authors: Yanxiong Li, Wenchang Cao, Jiaxin Tan, Qianqian Li, Guoqing Chen

Categories: eess.AS | 15 pages, 10 figures, 10 tables, accepted for publication in IEEE TASLP Score: 5.50/10 (Obj:8 Id:5 Ind:8 Comp:5 Eff:6 Nov:3 Rep:8)

  • Strength: 中 — 在 FCAC 的特定设置下确实取得了优于部分 baseline 的结果，且计算复杂度较低，对实际音频分类系统有一定参考价值。
  • Weakness: 论文的三个核心组件均有明确的先前工作：(1) Stochastic classifier for FSCIL: S3C (Kalla & Biswas, ECCV 2022) 已提出使用随机分类器（每类一个分布）进行 few-shot class-incremental learning，且包含均值和方差。本文的 stochastic classifier 在结构上与 S3C 高度相似。
  • Full review: Claude Code 全文七公理审稿

Few-shot Class-incremental Audio Classification (FCAC) aims to progressively recognize incremental classes with few tagged samples and meanwhile memorize base classes. To achieve satisfactory FCAC performance, the model needs to have high stability (memorizing base classes) and strong plasticity (adapting to incremental classes). In this work, we design a model which can be decoupled into two independent modules, namely an embedding learner and a stochastic classifier. The former is the backbone of a residual convolutional network, while the latter is composed of distributions and each distribution consists of a mean vector and a variance vector for representing one class. After being trained in the base session, the embedding learner is not updated in each incremental session and thus can memorize the knowledge of base classes. To make the embedding learner possess strong representation ability for incremental classes, we propose a strategy to pseudo-incrementally train the embedding learner using data augmentation in the base session. On the other hand, the stochastic classifier is continually updated in each incremental session and thus can adapt to incremental classes. Our model which consists of a pseudo-incrementally trained embedding learner and a continually updated stochastic classifier can increasingly identify incremental classes without forgetting base classes. Three datasets (FSC-89, NSynth-100 and LS-100) are used to verify the effectiveness of our method. Experiments show that our method exceeds the comparison methods in accuracy, and has lower complexity than most of the comparison methods. The code is at https://github.com/vinceasvp/PITEL-CUSC.

19. Audio Sentiment Analysis via Distillation and Cross-Modal Integration of Generated Multilingual Transcripts

Authors: Andrei-George Durdun, Victor Constantinescu, Radu Tudor Ionescu

Categories: cs.CL, cs.AI, cs.LG, cs.SD | Accepted at KES 2026 Score: 5.26/10 (Obj:6 Id:5 Ind:8 Comp:5 Eff:5 Nov:4 Rep:8)

  • Strength: 在 MSP-Podcast 大规模数据集上验证了 ASR+NMT 生成的多语言文本作为 privileged information 可为多模态 teacher 带来 +5.89% macro-F1 提升，且通过 KD 将增益部分转移至音频-only student（+1.54% macro-F1），推理零额外开销，代码开源。
  • Weakness: 核心方法（CCMT、KD/LUPI、ASR 文本增强）均为已有工作复用，与 Muaz et al. (2024) 高度重叠；多语言翻译的增量被自身消融证明有限（Audio+En vs. Audio+En+De+Fr 差异 <1.3%），且缺少与 INTERSPEECH 2025 Challenge top systems 及 label smoothing 正则化的公平比较。
  • Full review: Claude Code 全文七公理审稿

Automatically recognizing the sentiment, positive or negative, from speech is a challenging task, requiring both the analysis of vocal inflections and the interpretation of uttered words. Recent solutions rely on audio foundation models to solve the task, but it remains unclear if such models can take all aspects into account. To this end, we propose a multimodal solution that integrates audio and text information via cross-modal transformers, where text transcripts are automatically generated via an automatic speech recognition (ASR) tool. Moreover, we create multiple text modalities by automatically translating the transcripts into multiple languages via machine translation tools. Audio and multilingual text features are combined via a cascaded architecture comprising cross-modal transformer blocks that integrate modalities one by one. We further distill knowledge from the multimodal model, called teacher, into a unimodal (audio only) model, called student. We conduct experiments on a large-scale dataset, demonstrating that the automatically generated textual information can bring significant performance boosts in multimodal sentiment polarity classification. Our ablation study confirms that both automatic transcripts and automatic translations are helpful. Moreover, we show that the audio-only model can be enhanced via distillation, boosting performance without any computational overhead during inference. To reproduce the reported results, we publicly release our code at https://github.com/andreidurdun/cross-modal-audio-sentiment.

20. Multimodal Video-to-Music Recommendation via Semantic Retrieval and Temporal Reranking

Authors: Seungheon Doh, Minhee Lee, Sangmoon Lee, Ben Sangbae Chon, Juhan Nam

Categories: cs.MM, cs.SD | Accepted for publication at The Machine Learning for Audio workshop at ICML 2026 Score: 5.20/10 (Obj:6 Id:5 Ind:5 Comp:5 Eff:5 Nov:4 Rep:6)

  • Strength: 两阶段 retrieve-then-rerank 架构在 V2M 推荐上将 R@10 从最强 baseline 的 14.2 提升到 18.3（MedR 从 75 降至 46），且人类偏好研究中 Music Quality 以 96% win+tie 大幅超越生成式 baseline VidMuse。
  • Weakness: 核心提升可能主要来自 task-specific fine-tuning 而非方法创新（baselines 均为 zero-shot 通用模型），且绝对指标偏低（R@10=18.3 意味着超过 80% 的查询在 top-10 中找不到正确曲目），同时缺少组件级消融实验和跨数据集验证。
  • Full review: Claude Code 全文七公理审稿

We present VTMR, a two-stage framework for Video-To-Music Recommendation. In Stage~1, VTMR aligns comprehensive video and music signals in a joint audio-visual-text representation space and efficiently retrieves semantically compatible candidates using coarse global embeddings. In Stage~2, it reranks the retrieved candidates by attending to the temporal sequences of both video and music, thereby capturing fine-grained temporal correspondence. Evaluated on the video-to-music recommendation task, the multimodal retrieval stage improves R@10 from 14.2 to 15.9 and Median Rank from 75 to 58 over the strongest baseline; the temporal reranker further boosts R@10 to 18.3 and Median Rank to 46, demonstrating complementary gains from richer query encoding and temporal alignment. A human preference study confirms that VTMR is on par with a commercial baseline in overall preference, while outperforming a generative baseline in music quality.

21. TriA Pipeline: A Large-Scale Automatic Audio Annotation Pipeline For Audio Classification In Specific Scenarios

Authors: Hong Lyu, Mingru Yang, Qianhua He, Yanxiong Li, Jinxin Huang et al.

Categories: eess.AS, cs.AI, cs.LG | 5 pages, 2 figures, 4 tables, accepted for publication in Interspeech 2026. The code is at: https://github.com/huanxian/TriA Score: 4.79/10 (Obj:6 Id:5 Ind:5 Comp:3 Eff:6 Nov:3 Rep:8)

  • Strength: 低 — 四模块管线均为现成工具直接调用，无算法层面的贡献，无模块级消融，无设计原理说明。
  • Weakness: TriA Pipeline 是典型的工程组合：Standardization（格式标准化，标准做法）+ AAD（用 auditok 工具做能量阈值分割，现成工具）+ AED（用 BEATs 预训练模型做事件检测，直接调用）+ Filtering（用 audiobox-aesthetics 和 CLAP 做质量过滤，两个现成模型）。
  • Full review: Claude Code 全文七公理审稿

There are some datasets of varying scales for audio classification (AC) applied to different tasks. However, annotated data is limited for most scenarios, such as domestic environments. To address this challenge, we propose an $\textbf{A}$utomatic $\textbf{A}$udio $\textbf{A}$nnotation Pipeline–TriA Pipeline, which can efficiently convert audio from various scenarios into high-quality training data with audio event annotations. A TriA dataset was constructed with the TriA Pipeline, over 2130 hours of audio covering 431 audio classes. Furthermore, we partitioned a prior-knowledge-guided subset (TriA${\mathrm{GK}}$) from TriA and conduct comparative experiments on three domestic AC tasks. Comparing the result on manually annotated data only and that on manually annotated data combines TriA${\mathrm{GK}}$, TriA${\mathrm{GK}}$ could achieve average relative gains of 3.97% in accuracy and 3.35% in Macro-F1, validating the effectiveness of TriA${\mathrm{GK}}$ and the TriA Pipeline.

22. Learning-based Physics-Constrained Neural Kernel for Sound Field Estimation With Source-Position-Dependent Directional Weighting

Authors: Mattia Marella, Shoichi Koyama

Categories: cs.SD, eess.AS | Accepted to International Workshop on Acoustic Signal Enhancement (IWAENC) 2026 Score: 4.53/10 (Obj:6 Id:5 Ind:8 Comp:5 Eff:3 Nov:4 Rep:5)

  • Strength: 通过在方向加权 INR 中引入源位置依赖，将 per-snapshot 优化转为多源训练范式，在 75–525 Hz 范围内 NMSE 优于 snapshot-based 方法，方向加权函数可视化显示与真实源和一阶镜像源方向对齐。
  • Weakness: 仅在单一模拟房间、单一 T60、M=18 麦克风、窄频率范围内与同框架内两个 baseline 比较，未纳入领域内其他主流方法（NN-based、diffusion、sparse），实验设置无法支撑泛化 claim。
  • Full review: Claude Code 全文七公理审稿

A learning-based physics-constrained neural kernel for sound field estimation is proposed. Sound field estimation aims to estimate the spatial distribution of an acoustic field from a discrete set of microphone measurements, which have a wide range of applications. Among existing sound field estimation methods, kernel-regression-based methods offer a flexible and principled framework for incorporating physical constraints and allow inference through linear operation. It is also possible to adapt the kernel function to the target acoustic environment by representing the directional weighting function as an implicit neural representation (INR) and optimizing hyperparameters using measurements. However, the kernel function is generally optimized for single snapshot measurements of the microphones, which can lead to strong overfitting and poor generalization. We propose a source-position-dependent INR for the directional weighting function, enabling the kernel function to capture common directional patterns and to generalize to unseen source positions in the target acoustic environment. Experimental results indicate that our proposed method outperforms the snapshot-based method by estimating a directional weighting function that matches the directivity of the target sound field.

23. From Textural Counterpoint to Feature Encoding: A Multi-Dimensional Machine Representation Study of Haydn’s “The Lark” Integrating Electroacoustic Analysis

Authors: Yakun Liu, Zhiyu Jin, Hai Luan, Dong Liu, Xiaonan Li

Categories: cs.SD, cs.AI Score: 3.16/10 (Obj:5 Id:2 Ind:4 Comp:4 Eff:2 Nov:4 Rep:2)

  • Strength: 低 — 技术层面的”创新”（event-based timestamps、one-hot role encoding、特征拼接）均为已有技术的组合或简化版。真正的贡献在于跨学科叙事框架的构建，但这不构成可操作的方法贡献。
  • Weakness: 论文完全没有进行任何机器学习实验，因此无法识别任何有效原因
  • Full review: Claude Code 全文七公理审稿

Chamber music, as a highly precise multi-part interactive system, contains a logic of “role assignment and dynamic interaction” that provides an extremely valuable blueprint for exploring human-computer collaborative composition paradigms. Addressing the lack of role perception capabilities in existing deep music generation models during polyphonic interactions, this paper conducts an interdisciplinary analysis of Haydn’s String Quartet in D Major, The Lark (Op. 64, No. 5). We propose a novel research path: “Classical Morphology Qualitative Analysis-Electroacoustic Quantitative Measurement-Machine Representation Reconstruction.” The study first utilizes auditory analysis to dissect the counterpoint morphology of the leading voice and the underlying groove in the first movement. Subsequently, it introduces spectrum and dynamic feature analysis tools from a Digital Audio Workstation (DAW) to translate subjective auditory perception into objective, measurable physical parameters. Building on this, the paper introduces a fundamentally new approach to low-level computer feature extraction: completely abandoning the traditional mechanical quantization grid, introducing Event-based Timestamps to record the duration of micro-timing, and transforming acoustic features into an independent “Role-Aware Encoding” as an aesthetic heuristic mechanism (a phenomenological anchor). This study not only completes the logical loop spanning classical analysis, electronic music mapping, and AI symbolic generation but also establishes a profound theoretical foundation-from the perspectives of interactive aesthetics and media philosophy-for constructing human-computer collaborative music systems imbued with “social attributes” and “otherness awareness.”

24. Designing Maintainable Hybrid Generative Systems: A Quantum-Inspired Approach to Automated Music Harmony Generation

Authors: Josef Pavlicek

Categories: cs.SD, cs.AI, cs.ET, eess.AS | 12 pages, 1 figure, 4 tables. Extended version of the 4-page paper accepted at the 34th International Conference on Information Systems Development (ISD2026, Prague). Source code and dataset available at https://github.com Score: 3.05/10 (Obj:5 Id:2 Ind:2 Comp:4 Eff:2 Nov:2 Rep:7)

  • Strength: 系统架构模块化清晰（生成器+规则后处理），代码和数据公开可复现，无需训练语料即 可生成结构和声
  • Weakness: 无任何外部 baseline 对比（仅自比较），优化层的改善是规则定义的同义反复，11首 C 大调旋律无法支撑泛化，核心框架来自作者一天前发表的前置工作 (arXiv:2607.05007)
  • Full review: Claude Code 全文七公理审稿

This paper presents the design and evaluation of a maintainable hybrid generative architecture for automated music harmony generation from melody. The proposed system combines quantum-inspired candidate exploration over overlapping melodic contexts with explicit rule-based optimization to balance generative flexibility and structural control. The architecture is evaluated using explicit and reproducible metrics covering structural coherence, functional agreement, harmonic similarity, and robustness. The results show that the proposed approach produces harmonizations that preserve tonal structure and cadential behavior while allowing multiple valid harmonic realizations. Furthermore, the optimization layer improves structural coherence, stability, and predictability without requiring a training corpus. The study demonstrates that transparent and controllable hybrid generative systems can be systematically designed and evaluated within the context of Information Systems Development.