论文解读
Sidon: Fast and Robust Open-Source Multilingual Speech Restoration for Large-Scale Dataset Cleansing
语音增强 | 8.5/10
开源工具
论文解读
The Singing Voice Conversion Challenge 2025: From Singer Identity Conversion to Singing Style Conversion
歌唱语音转换 | 7.0/10
论文解读
Via Score to Performance: Efficient Human-Controllable Long Song Generation with Bar-Level Symbolic Notation
音乐生成 | 7.5/10
论文解读
Z-Scores: A Metric for Linguistically Assessing Disfluency Removal
模型评估 | 6.5/10
论文解读
Opening the Design Space: Two Years of Performance with Intelligent Musical Instruments
音乐生成 | 6.5/10
论文解读
Audio Video Verbal Analysis (AVVA) for Capturing Classroom Dialogues
音频问答 | 6.0/10
论文解读
TTS-PRISM: A Perceptual Reasoning and Interpretable Speech Model for Fine-Grained Diagnosis
语音质量评估 | 7.5/10
每日研究速递
语音/音乐/音频论文速递 2026-04-27
共分析 13 篇语音/AI 论文
论文解读
DiariZen Explained: A Tutorial for the Open Source State-of-the-Art Speaker Diarization Pipeline
说话人分离 | 6.5/10
论文解读
Centering Ecological Goals in Automated Identification of Individual Animals
这篇论文旨在解决一个关键问题：为什么近年来在动物个体自动识别（基于图像或声音）上报告的高准确率算法，却很少转化为生态学实践中的常规工具？其方法核心是提出一个“以生态目标为中心”的评估与部署框架，强调自动化识别的有用性取决于其服务的具体生态问题、可用数据以及错误类型带来的实际后果。与以往主要关注算法准
论文解读
Reducing the Offline-Streaming Gap for Unified ASR Transducer with Consistency Regularization
本文旨在解决训练单一自动语音识别（ASR）模型同时高效支持高精度离线转写和低延迟流式识别这一挑战。现有统一模型在低延迟流式模式下性能下降明显。作者提出了一个统一的RNN-Transducer (RNNT) 框架，其核心是结合了**带右上下文的chunk限制注意力**和**动态chunk卷积（DCCo