论文解读
TextlessRAG: End-to-End Visual Document RAG by Speech without Text
语音问答 | 8.5/10
端到端
论文解读
Tokenchain: A Discrete Speech Chain via Semantic Token Modeling
语音识别 | 7.0/10
论文解读
Toward Robust And Efficient Beat Tracking Via Beat-Aware Attention
音乐理解 | 8.5/10
论文解读
Tpeformer: Temporal Patch Embedding Transformer
语音情感识别 | 7.5/10
论文解读
Train Short, Infer Long: Speech-LLM Enables Zero-Shot Streamable Joint ASR and Diarization on Long Audio
说话人分离 | 9.0/10
论文解读
Tri-Attention Fusion: Joint Temporal-Spectral and Bidirectional Modeling for Speech Spoofing Detection
语音伪造检测 | 7.0/10
论文解读
UJCodec: An End-to-end Unet-Style Codec for Joint Speech Compression and Enhancement
语音增强 | 7.5/10
论文解读
UMA-SPLIT: Unimodal Aggregation for Both English and Mandarin Non-Autoregressive Speech Recognition
语音识别 | 7.5/10
论文解读
USVexplorer: Robust Detection of Ultrasonic Vocalizations with Cross Species Generalization
音频事件检测 | 8.0/10
论文解读
VBx for End-to-End Neural and Clustering-Based Diarization
说话人分离 | 8.5/10
论文解读
VChangeCodec: An Ultra Low-Complexity Neural Speech Codec with Built-In Voice Changer for Customized Real-Time Communication
语音转换 语音增强 | 8.0/10
论文解读
WhisperPipe: A Resource-Efficient Streaming Architecture for Real-Time Automatic Speech Recognition
语音识别 | 6.5/10
论文解读
β-AVSDNET: A Novel End-To-End Neural Network Architecture For Audio-Visual Speaker Diarization
说话人分离 | 7.5/10
论文解读
Full-Duplex Interaction in Spoken Dialogue Systems: A Comprehensive Study from the ICASSP 2026 HumDial Challenge
语音对话系统 | 6.5/10
论文解读
Reducing the Offline-Streaming Gap for Unified ASR Transducer with Consistency Regularization
1. **问题**：训练一个既能高精度离线转录又能低延迟流式识别的统一ASR模型极具挑战性，传统方法在低延迟下性能会急剧下降。 2. **方法核心**：提出一个统一的Transducer框架，结合分块注意力（含右上下文）和动态块卷积（DCConv）来适配两种模式。核心创新是引入了模式一致性正则化损失
论文解读
Deep Supervised Contrastive Learning of Pitch Contours for Robust Pitch Accent Classification in Seoul Korean
这篇论文旨在解决将连续变化的基频（F0）曲线映射到首尔韩语中离散、不变的音高重音类别（如LHLH, HHLH）这一难题。传统方法易受F0测量噪声和说话人差异的影响。为此，作者提出了**Dual-Glob**，一个深度监督对比学习框架。其核心是通过一个**双分支（干净视图和增强视图）编码器**，在共享
论文解读
Text-To-Speech with Chain-of-Details: modeling temporal dynamics in speech generation
本文针对文本转语音（TTS）任务，提出了一种名为“细节链”（Chain-of-Details, CoD）的新框架。**要解决的问题**是现有TTS方法在建模语音生成的时域动态（从粗略时序到精细声学细节的渐进过程）方面存在不足。**使用的方法**是将语音生成分解为多个时间分辨率递增的阶段，在每个阶段使
论文解读
MUSCAT: MUltilingual, SCientific ConversATion Benchmark
本文提出了 MUSCAT，一个用于评估多语言科学对话场景下自动语音识别（ASR）性能的新基准。数据集包含 6 组双语对话录音（共约 65 分钟，9,066 词），涉及英语与德语、土耳其语、中文、越南语的配对对话；每组对话使用 Meeting Owl 3、ReSpeaker USB 麦克风阵列和 Me
论文解读
VoxMind: An End-to-End Agentic Spoken Dialogue System
端到端语音对话模型在自然交互上进步迅速，但普遍缺乏处理复杂任务的agent能力（工具调用、规划、推理）。本文首先形式化定义了"端到端语音智能体"的四大维度——画像（Profile）、记忆（Memory）、规划（Planning）与执行（Action Execution），填补了该领域理论标准的空白。
论文解读
An Ultra-Low Latency, End-to-End Streaming Speech Synthesis Architecture via Block-Wise Generation and Depth-Wise Codec Decoding
这篇论文旨在解决实时交互式语音合成中**推理延迟高**与**声学质量（尤其是高频细节）易受损**的核心矛盾。传统流水线依赖计算密集的神经声码器进行波形重建，且基于连续回归的声学模型易导致频谱过平滑。为