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。

2026-07-20
日期2026-07-20
已评分—
均分—
最高—

Daily Papers — 2026-07-20

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

1. Dense-Sparse Dynamic Time Warping for Customizing Piano Concerto Accompaniments

Authors: TJ Tsai, Kavi Dey, Yigitcan Ozer, Meinard Muller

Categories: cs.SD, eess.AS | Published at ICASSP 2025 Score: 6.70/10 (Obj:6 Id:8 Ind:5 Comp:8 Eff:5 Nov:8 Rep:8)

  • Strength: DS-DTW 用单一超参数 γ=0.8、无可训练参数，在 2s tolerance 下将 alignment error 从 2.7% 降至 0.6%（78% 相对降低），达到与训练过的 HDemucs 分离模型可比水平。
  • Weakness: Benchmark 仅 4 个 concerto movements、1 位 pianist、靠 TSM 扩充到 150 scenarios；Puser 在录制时已对齐 Oacc，核心评测的独立性存在构造耦合；0.1s tolerance 下 53% error rate 表明精确对齐远未解决。
  • Full review: Claude Code 全文七公理审稿

In this study, we explore how pianists can customize Music Minus One (MMO) concerto accompaniments to match their playing style. Bypassing the need for a symbolic score, often not available digitally, we use three types of audio data: solo piano recordings, MMO orchestra-only recordings, and mixed recordings of both piano and orchestra (e.g., from YouTube). The mixed recording serves as an intermediary reference to align the solo and orchestra parts, with only the orchestral part being adjusted through time-scale modification to synchronize with the user’s playing. The main challenge with estimating these alignments is the spectral mismatch between recordings containing different musical parts. Motivated by this application scenario, we introduce Dense-Sparse DTW, a variant of Dynamic Time Warping (DTW) that is designed to improve robustness of alignments to spectral mismatch by focusing on aligning a selected subset of audio frames containing prominent timing cues. We collect and annotate data from four piano concerto movements and establish a framework for generating and evaluating customized accompaniment recordings. On this benchmark, we show that Dense-Sparse DTW has better or comparable performance than more complex approaches based on source separation and spectral subtraction techniques.

2. ESCUCHA: A Spanish Speech Benchmark for Heterogeneous Acoustic Conditions

Authors: Fernando López, Ana Ayala, Guillermo Segovia, Fernando Ibáñez, Ana Martínez et al.

Categories: cs.CL | Under review Score: 6.30/10 (Obj:7 Id:6 Ind:6 Comp:8 Eff:6 Nov:5 Rep:8)

  • Strength: 首个覆盖西班牙语 + 非规范性病理语音 (ALS/中风, 158题) 的 in-the-wild LALM benchmark，162.9 小时人工策展音频，全开源 (CC BY 4.0, GitHub repo 含数据+评分脚本+复现 notebook)，AIF 用确定性 verifier 无 judge 污染，cascade baseline (60%) 超过大多数端到端模型揭示了 benchmark 的词汇可恢复性。
  • Weakness: 1000 题规模偏小且无 inter-annotator agreement 报告（人工上限仅 1 人），分类法直接继承 MMAU-Pro 无原创重构，PolySpeech-100 (concurrent, 2026-05-31) 已含西班牙语语义推理削弱 “first” 声明，Gemini-2.5-Flash 因 payload 限制未在完整集评测导致闭源 SOTA 不可知，无 ablation 证明多音频/AIF/spoken 各组。
  • Full review: Claude Code 全文七公理审稿

As large audio language models (LALMs) advance, robust evaluation frameworks have become essential. In this context, Spanish speech understanding under realistic acoustic conditions has received particularly little attention. We introduce ESCUCHA, the first Spanish speech understanding benchmark designed to evaluate LALMs across heterogeneous acoustic conditions and reasoning abilities. ESCUCHA comprises 1,000 human-curated questions paired with audio, totaling 162.9 hours sourced directly ``from the wild’’ rather than drawn from existing datasets, with durations ranging from a few seconds to over 80 minutes. The benchmark emphasizes reasoning, spanning 9 perceptual and 10 reasoning categories, and it captures linguistic diversity through multiple Spanish accents and non-normative speech. ESCUCHA further includes multi-audio questions, spoken questions, and audio instructions, and it flags which questions support open-ended evaluation. Benchmarking several state-of-the-art multimodal and speech models reveals substantial performance gaps relative to trained humans.

3. When to Use Extra Context: Evidence-Grounded Terminology Adaptation for Simultaneous Speech Translation

Authors: Zeyu Yang, Satoshi Nakamura

Categories: cs.CL Score: 5.68/10 (Obj:6 Id:8 Ind:5 Comp:6 Eff:5 Nov:5 Rep:5)

  • Strength: EGTA 用确定性表面匹配做 evidence-conditioned 选择，在 MCIF-dev 上 NERclean recall +0.212/+0.266、acronym recall +0.099/+0.171（p<0.001），shuffled-memory 控制将增益压到 +0.012/+7.5，证明改进来自正确 paper memory 而非 generic biasing。
  • Weakness: 所有对比均为 self-baseline（同一 Qwen3-Omni LoRA 的不同 inference 配置），未与任何已发表的 IWSLT 2025 公开系统直接对比；术语记忆由 Qwen3-30B 提取、backbone 为同族 Qwen3-Omni，与 NERclean 评测目标存在弱闭环；无 code/checkpoint release 声明。
  • Full review: Claude Code 全文七公理审稿

Extra context is valuable for simultaneous speech translation of technical talks, but injecting the entire document context into every streaming segment is often too coarse. Through diagnostic experiments, we find that context gains mainly come from paper-specific terminology recovery rather than uniform semantic enhancement. We therefore propose EGTA, an Evidence-Grounded Terminology Adaptation framework that builds a document terminology memory, selects compact candidate terms conditioned on the current streaming state, and adapts ASR/speech-side and decoder-side decision spaces using only the selected terms. EGTA can be instantiated in cascaded, end-to-end, and generation-only SimulST settings without full-model fine-tuning. We evaluate EGTA on an ACL technical-talk SimulST evaluation suite consisting of MCIF-dev and ACL60/60-dev. On MCIF-dev, EGTA-RG improves BLEU by +1.05/+0.59, XCOMET-XL by +0.019/+0.006, named-entity recall by +79\%/+73\% relative, and acronym recall by +0.099/+0.171 on En$\rightarrow$Zh and En$\rightarrow$De. Across MCIF-dev latency settings, EGTA consistently improves XCOMET-XL, named-entity recall, and acronym recall. External validation on ACL60/60-dev further shows consistent terminology-recall gains without additional fine-tuning. Shuffled-memory controls and activation audits provide evidence that the improvements are tied to paper-specific evidence alignment rather than generic context prompting.

4. SpEmoC: A Balanced Speaker-Segment Multimodal Emotion Benchmark

Authors: Sania Bano, Shahzad Ahmad, Santosh Kumar Vipparthi, Sukalpa Chanda, Subrahmanyam Murala

Categories: cs.CV Score: 5.47/10 (Obj:8 Id:5 Ind:5 Comp:5 Eff:6 Nov:5 Rep:5)

  • Strength: 30k 类别平衡、movie-level split、三模态对齐的 MER 数据集，用 5 个 SOTA 模型（含 CVPR 2025 EMOE）验证 balanced supervision 使 minority class（Fear/ Disgust）F1 从近 0 提升至 36-71，imbalance gap ∆ 从 15-27 降至 ≤0.65。
  • Weakness: 因果识别不足（同时改变 balance/ scale/ split/ source 无隔离 ablation），标注模型参与数据筛选构成选择偏差闭环（50k→30k 依赖模型 logits 阈值），cross-dataset 迁移到 Neutral-dominant 数据集时 W-F1 仅 17-33，实用价值受限。
  • Full review: Claude Code 全文七公理审稿

Understanding human emotions in spoken conversations is a key challenge in affective computing, with applications in empathetic AI, human computer interaction, and mental health monitoring. However, existing datasets vary in scale, emotion distribution, modality alignment, and data partitioning strategies, which can influence reliable cross-dataset generalization and minority-emotion modeling. We introduce SpEmoC a Speaking segment Emotion for Conversations comprising 306,544 raw clips from 3,100 English language movies and TV series. From these, 30,000 high quality, class balanced clips are curated, featuring synchronized visual, audio, and textual modalities annotated for seven emotions through a hybrid pipeline that integrates pretrained models with human validation. SpEmoC uses strict movie- and series-level splits to prevent content overlap between split sets, allowing more reliable evaluation of model generalization. The dataset also maintains a near-balanced distribution across seven emotions, including minority classes such as Fear and Disgust, which supports more balanced learning across categories. Extensive experiments, including in-domain benchmarking, cross-dataset transfer, low-data training, class-imbalance analysis, and modality transfer show that balanced data and careful splitting lead to more stable performance across emotions when models are evaluated on other datasets. These results highlight the importance of dataset design for robust and transferable multimodal emotion recognition.

5. FillGauss: Fine-Grained Filling-Aware Impact Sound Generation for 3D Gaussian Splatting

Authors: Chen Yang, Ganye Wen, Bin Huang, Jiayi Lyu, Zehai Niu et al.

Categories: cs.MM Score: 5.37/10 (Obj:8 Id:5 Ind:5 Comp:5 Eff:5 Nov:5 Rep:6)

  • Strength: 提出首个 filling-aware 冲击声任务与数据集 FillImpact（88 物体/5000+ 录音），数据物理一致性经 f∝√(k/m) 等经典声学定律验证，fill material/level 属性盲测匹配率 52-60% 显著优于 baseline。
  • Weakness: 主指标 FAD=1.3500 输给自身 backbone TangoFlux (FT) 的 1.0627（+27%），3DGS 分支可能引入噪声；方法本质是 SonicGauss+文本 prompt+TangoFlux 拼装，仅 2 个 baseline，代码/数据集未明确开源。
  • Full review: Claude Code 全文七公理审稿

Synthesizing physically plausible impact sounds from visual observations remains a great challenge in multi-modal AI. Existing 3D-aware audio generation methods primarily model the surface geometry of hollow rigid bodies. However, they fundamentally overlook internal filling states, a critical physical factor that drastically modulates acoustic resonance and damping. To address this issue, we have defined a new task called Fine-Grained Filling-Aware Impact Sound Generation. As a foundational step, we first introduce the fine-grained fill-aware dataset (FillImpact), a pioneering multi-modal collection comprising over 5,000 rigorous acoustic recordings from 88 diverse real-world objects. It captures impact interactions with varying internal contents (i.e., water, rice), a continuous range of fill levels, and distinct striker materials. Furthermore, comprehensive acoustic analysis confirms that the collected data closely aligns with established physical laws governing acoustic resonance and damping, indicating its suitability for physically grounded modeling. Building on this dataset, we propose a novel generative framework (FillGauss) that integrates 3D Gaussian Splatting (3DGS) with internal state conditioning for sound generation. By fusing 3DGS geometric features, precise 3D spatial strike coordinates, and fine-grained textual physical conditions within a latent diffusion architecture, FillGauss enables position-aware, striker-aware, and filling-aware audio generation. Extensive experiments demonstrate that our approach could generate high-fidelity impact sounds that adhere to underlying physical principles, establishing a new state-of-the-art for physically grounded cross-modal audio generation.

6. X-Translator: A Real-Time Multilingual Speaker-Aware Speech-to-Speech Translation System

Authors: Yuxiang Zhao, Yichi Zhang, Yanjie An, Yanqiao Zhu, Zhanxun Liu et al.

Categories: eess.AS Score: 5.26/10 (Obj:8 Id:5 Ind:5 Comp:5 Eff:5 Nov:4 Rep:8)

  • Strength: 模块化级联式 S2ST 在 speaker preservation 上清晰取胜（短形式 Spk. SIM 0.87/0.88 第一，多说话人 0.77/0.77 并列第一），代码/配置/评估全开源，运行时日志可追踪每个模块延迟与 prompt 路由错误。
  • Weakness: 全文无任何消融实验隔离 commitment layer 或 speaker prompt manager 的贡献，无法将 Spk. SIM 优势归因于 runtime design 而非 X-Voice backend；延迟全部系统中最慢（Start 2584/4056 ms），翻译质量 COMET 在两方向均落后于至少一个闭源 API。
  • Full review: Claude Code 全文七公理审稿

Real-time speech-to-speech translation (S2ST) systems must balance translation quality, latency, speech naturalness, and speaker consistency. Publicly documented S2ST systems have advanced direct, multilingual, streaming, and expressive modeling, while proprietary products and APIs increasingly expose real-time translation capabilities to users. However, practical deployment remains challenging for open and reproducible systems, especially in long-form and multi-speaker conversations where partial ASR hypotheses are unstable, turn boundaries are ambiguous, and target speech must be generated with an appropriate speaker prompt. We present X-Translator, a low-cost modular cascaded S2ST system that combines streaming ASR, machine translation, and prompt-conditioned TTS through a session-level runtime controller. The system uses incremental segment commitment to convert unstable ASR streams into translation-ready units, and an online speaker prompt manager to bind source speech spans to speaker-specific voice prompts for synthesis. We evaluate translation, speech quality, and latency with OpenSTBench, compare against proprietary speech translation APIs as behavioral baselines, measure long-form voice stability, evaluate speaker preservation in multi-speaker conversations, and assess multilingual translation quality. X-Translator provides an open platform for understanding the practical trade-offs of deployment-oriented S2ST. Code and demo are available at https://github.com/zhaoyx239/X-Translator.

7. Time-Frequency Consistency Learning for Robust Speech Deepfake Detection

Authors: Jun Xue, Zhuolin Yi, Yanzhen Ren, Yihuan Huang, Jiayu Xiong et al.

Categories: cs.SD, cs.AI | Accepted by ACM MM 2026 Score: 5.20/10 (Obj:6 Id:5 Ind:5 Comp:5 Eff:5 Nov:4 Rep:8)

  • Strength: TFCL 在 AFE pipeline 最严苛条件 (VAD) 下将 EER 从 22.20% 降至 9.78%（vs XLSRAASIST baseline，56% 相对降低），且跨三个 backbone (AASIST/Nes2Net/MultiConv) 一致有效，训练时双输入设计推理时无额外开销。
  • Weakness: 高层方法范式（一致性学习 for 通信鲁棒 SDD）已被 RTCFake (ACL 2026, 3 个月前提交) 提出 but 未被引用；仅单一数据集 + 模拟 AFE 验证，缺少真实平台和跨数据集证据；VAD 下 9.78% EER 对安全部署仍偏高。
  • Full review: Claude Code 全文七公理审稿

Recently, speech deepfake detection (SDD) has achieved significant progress. However, its robustness evaluation remains largely confined to controlled additive noise scenarios, lacking systematic investigation of the complex distortions introduced by acoustic front-end (AFE) processing pipelines in real-world deployments. In this work, we simulate a unified AFE pipeline comprising acoustic echo cancellation, noise suppression, automatic gain control, and voice activity detection (VAD), and conduct a comprehensive evaluation of current state-of-the-art models. The results show that the nonlinear and time-frequency coupled distortions introduced by AFE significantly degrade detection performance. To address this issue, we propose a Time-Frequency Consistency Learning (TFCL) framework, which aims to learn invariant spoofing representations that remain stable before and after AFE processing. We observe that AFE not only introduces temporal misalignment (e.g., segment-level shifts caused by VAD), but also weakens or distorts critical frequency-domain cues. To this end, TFCL employs an attention-driven soft alignment mechanism to capture cross-temporal dependencies, along with frequency-domain structural consistency constraints to enforce feature invariance. As a result, the model is able to maintain stable representations under both temporal perturbations and spectral distortions. Extensive experimental results demonstrate that the proposed method effectively mitigates the performance degradation caused by AFE processing, significantly improving the robustness of SDD in real-world scenarios. The code is available at https://github.com/JunXue-tech/TFCL.

8. SSTMark: Robust Training-Free Semantic-Level Speech Watermarking

Authors: Kuan-Lin Chu, Jun-Cheng Chen, Chun-Shien Lu

Categories: cs.SD Score: 4.74/10 (Obj:8 Id:5 Ind:5 Comp:5 Eff:6 Nov:4 Rep:3)

  • Strength: 通过 STT→文本水印→TTS 的 semantic-level pipeline 在 AudioMarkBench 上实现最稳定的平均鲁棒性（avg TPR 90.2% signal-processing / 90.0% codec），且对抗攻击 HSJA 需 SI-SNR=2.688 远低于 AudioSeal 的 48.420。
  • Weakness: 无攻击时 TPR 仅 91.0%（所有 baseline 为 100%），方法本质是 Whisper+Yang2023+CosyVoice 三模块串联无新算法组件，无代码开源且未与同期 LambdaMark/DuraMark 比较。
  • Full review: Claude Code 全文七公理审稿

As speech generation models become increasingly realistic and widely accessible, concerns about the misuse, attribution, and governance of synthetic speech continue to grow. Watermarking provides a practical way to make synthesized speech traceable and verifiable. Most existing speech watermarking methods embed watermark information into signal-level representations, such as waveforms or spectrograms. Under sufficiently strong distortions, the embedded watermark may be weakened or destroyed, leading to degraded detectability. In this paper, we propose SSTMark, a training-free speech watermarking framework that operates at the semantic level through text watermarking. Unlike conventional signal-level watermarking methods, SSTMark encodes watermark information into the semantic content conveyed by generated speech, and detects the watermark from the recovered linguistic content. Experiments on AudioMarkBench demonstrate that SSTMark exhibits the strongest average robustness. Compared with the state-of-the-art baselines at a fixed false positive rate of 1\%, SSTMark improves the average detection rate by 4.6\% and 16.9\% on signal-processing edits and compression edits, respectively.

9. Harness TTS: Towards Context-Aware Expressive Speech Synthesis with Harness Layer

Authors: Shengfan Shen, Di Wu, Xingchen Song, Dinghao Zhou, Pengyu Cheng et al.

Categories: cs.SD Score: 4.53/10 (Obj:5 Id:4 Ind:2 Comp:5 Eff:5 Nov:5 Rep:4)

  • Strength: 将 TTS 风格控制重构为闭集 prompt-tool routing，4B planner 在 explicit 子集达 74.3% Top-1 且 P95 First-ID 仅 41.2ms，在 CosyVoice3 上 instruction-following win rate 领先 Instruct 基线 23.1–35.6 个百分点，工程可部署性清晰。
  • Weakness: 测试数据、routing ground-truth label、合成 win-rate judge 三者全部来自 Gemini 闭环且无独立人类校验，implicit 子集 Top-1 仅 43.0% 表明上下文感知能力本身薄弱，同时缺 random/oracle baseline 及对 CapTalk 等直接同类的对比，核心结论的认识论效力不足。
  • Full review: Claude Code 全文七公理审稿

Expressive speech synthesis for voice assistants requires flexible style control that adapts to explicit requests and broader interaction context. We propose Harness TTS, a lightweight control layer that wraps around a TTS engine to externalize and govern its expressive behavior. It reformulates style control as closed-set prompt-tool routing: offline, a compact registry of stylistic prompt tools is constructed with structured metadata; online, an LLM planner selects the appropriate tool based on a priority-aware observation schema, and the TTS executor synthesizes speech using the corresponding prompt audio. We evaluate Harness TTS on both routing and synthesis tasks. In routing, Qwen3-4B achieves Top-1 accuracies of 74.3%, 43.0%, and 64.6% on explicit, implicit, and conflict subsets. For synthesis, experiments on CosyVoice3 and VoxCPM2 show that Harness TTS outperforms instruction-only control, achieving higher instruction-following win rates (margins of 23.1-35.6 points on CosyVoice3 and 13.8-20.0 points on VoxCPM2) and improving UTMOSv2 scores by 0.11-0.38. Moreover, the 4B planner delivers its first tool recommendation in under 50 ms in standard mode, introducing negligible latency for real-time interaction. These results demonstrate that equipping TTS engines with a dedicated Harness layer offers a practical, auditable, and context-aware solution for voice assistant expression control.

10. FlowSonic: Stable Zero-Shot Music Editing via High-Order Trajectory Integration

Authors: Ali Boudaghi, Hadi Zare

Categories: cs.SD, cs.LG, eess.AS, eess.SP, math.NA Score: 4.05/10 (Obj:8 Id:5 Ind:6 Comp:5 Eff:4 Nov:3 Rep:8)

  • Strength: DHC 初始化机制在 genre-transfer 上将 FluxMusic 编辑 MOS 从 1.00 提升至 3.23，并在消融中展示 AB3+DHC 相对 Heun 的频谱稳定性优势（Table I-II）。
  • Weakness: 论文未引用同作者前作 MusRec (arXiv 2511.04376) 与同期 ABM-RF-editing 工作 (arXiv 2503.16522)，并以”无公开代码”为由排除 MelodyFlow/SteerMusic，但二者代码与权重实测均已开源，导致新颖性与 baseline 公平性双重失效。
  • Full review: Claude Code 全文七公理审稿

Zero-shot text-guided editing of real-world music recordings requires balancing semantic modification with faithful preservation of the original musical structure. Although recent diffusion transformers trained with rectified flow have achieved remarkable success in text-to-music generation, extending them to edit existing recordings remains challenging because editing requires accurate deterministic inversion, reliable structural preservation, and numerically stable integration throughout the inversion and generation processes. We present FlowSonic, a zero-shot music editing framework built upon a pretrained diffusion transformer trained with rectified flow. FlowSonic first deterministically inverts a real-world recording into the latent space and preserves its musical structure during editing by reusing cross-attention representations extracted during inversion. To improve the numerical reliability of inversion-based editing, we introduce a high-order ODE solver and systematically investigate how different numerical integration schemes influence trajectory stability, structural preservation, and semantic controllability. Comprehensive experiments on timbre-transfer and genre-modification tasks demonstrate that FlowSonic consistently outperforms existing music editing methods across semantic alignment, harmonic preservation, structural consistency, and perceptual audio quality. We further provide geometric and empirical analyses showing how the proposed numerical integration strategy improves latent trajectory stability and leads to more reliable music editing.

11. Audio Cross Verification Using Dual Alignment Likelihood Ratio Test

Authors: Heidi Lei, Arm Wonghirundacha, Irmak Bukey, TJ Tsai

Categories: cs.SD | Published at ICASSP 2023 Score: 3.79/10 (Obj:5 Id:2 Ind:5 Comp:5 Eff:5 Nov:2 Rep:9)

  • Strength: 无学习、可解释的 DA-LRT 在 DAPS 合成 benchmark 上对 insertion/deletion（L≥0.5s）达到 EER≤0.4%，且代码开源、运行时 <6 秒/查询。
  • Weakness: 仅与极弱 MFCC-Euclidean baseline 比较，完全遗漏同一团队 2020 年 SNWTW+BiLSTM（EER 0.43%）这一最直接强 baseline，无法识别真实增益来源；replacement 短编辑（L=0.25s）EER 高达 33-38% 不可用。
  • Full review: Claude Code 全文七公理审稿

This paper explores a way to verify that audio has not been maliciously tampered in a specific context: short viral videos taken from news recordings. Rather than trying to detect artifacts of tampering (internal inconsistency), we focus on positively verifying a query against a trusted source such as a news recording (external consistency). We propose a method for cross verifying a short audio query against a reference recording from which it was taken. Our approach is to define two hypotheses (non-tampered vs tampered), calculate the most likely alignment between query and reference for each hypothesis, and then perform a likelihood ratio test on the two alignments. We show that this method is fast to compute, much more robust than using MFCC features with Euclidean distance, and has the key benefit of explainability. Our cross verification approach provides an alternative perspective and complementary tool to existing tampering detection methods.

12. The tttAI System for the TSA-ASR Task of the SmartGlasses Challenge 2026

Authors: Xuanji He, Gaoyang Dong, Xiaoxiao Li, Minchuan Chen, Fengjie Zhu

Categories: eess.AS Score: 3.79/10 (Obj:6 Id:2 Ind:6 Comp:2 Eff:5 Nov:2 Rep:5)

  • Strength: 在 SmartGlasses Challenge 2026 Track 2 多人会议场景取得 tcpCER 34.04%、排名 leaderboard 第 2，dominant-speaker fallback 策略在 dev set 上将 tcpCER 从 24.69% 降至 22.88%（-1.81 点）。
  • Weakness: 系统全部由已发表工具包（DiariZen/WeSep/FireRedASR2-AED）串联构成，无新机制或因果识别实验，所有 ablation 共享 dev set 且未与挑战赛第 1 名或同 backbone 强 diarization 基线公平对比。
  • Full review: Claude Code 全文七公理审稿

This paper presents the tttAI system submitted to the TSA-ASR task of the SmartGlasses Challenge 2026, evaluated on both two-person dialogues (Track 1) and multi-party meetings (Track 2). The task requires time-stamped speaker-attributed speech recognition from smart-glasses recordings. This is particularly challenging due to long-form audio, multiple speakers, and frequent overlapping speech. We proposed a cascaded architecture consisting of speaker diarization, overlap detection, target-speaker extraction, post-processing, and automatic speech recognition. The diarization module extracts features via WavLM-Large, performs frame-wise speaker classification with a Conformer encoder, and then generates global speaker segments through embedding clustering. For overlapped regions, we apply a WeSep-based target-speaker extraction model with ECAPA-TDNN speaker embeddings. When the extraction is unreliable, a dominant-speaker fallback strategy is used. The final system uses FireRedASR2-AED with the first microphone channel. The submitted system has a total parameter count of approximately 1.53B. On Track 1, our system achieves a tcpCER of 7.10%. On Track 2, it achieves a tcpCER of 34.04% and ranks second on the leaderboard.