论文解读
Towards Understanding Modality Interaction in Multimodal Language Models via Partial Information Decomposition
音视频理解 | 5.3/10
可解释性
论文解读
Decomposer: Learning to Decompile Symbolic Music to Programs
音乐理解 | 8.4/10
每日研究速递
语音/音乐/音频论文速递 2026-07-03
共分析 31 篇语音/AI 论文
论文解读
Automatic Detection of Stress from Speech in the Trier Social Stress Test
语音情感识别 | 7.4/10
每日研究速递
语音/音乐/音频论文速递 2026-07-02
共分析 16 篇语音/AI 论文
论文解读
Profiling the Voice: Speaker-Specific Phoneme Fingerprinting for Speech Deepfake Detection
语音伪造检测 | 7/10
论文解读
Preliminary Insights in Chronos Frequency Data Understanding and Reconstruction
模型评估 | 6.0/10
论文解读
Discovering and Steering Interpretable Concepts in Large Generative Music Models
音乐生成 | 8.0/10
论文解读
Steering Autoregressive Music Generation with Recursive Feature Machines
音乐生成 | 8.0/10
论文解读
The Deleuzian Representation Hypothesis
模型评估 | 7.5/10
论文解读
Do Sparse Autoencoders Capture Concept Manifolds?
可解释性 | 7.0/10
每日研究速递
语音/音乐/音频论文速递 2026-05-01
共分析 21 篇语音/AI 论文
论文解读
Connecting Layer-Wise Representation of Wavlm with Spectro-Temporal Modulation on Speaker Verification
说话人验证 | 6.0/10
论文解读
GLoRIA: Gated Low-Rank Interpretable Adaptation for Dialectal ASR
语音识别 | 8.0/10
论文解读
ICLAD: In-Context Learning with Comparison-Guidance for Audio Deepfake Detection
本文针对音频深度伪造检测模型在真实场景（in-the-wild）中泛化能力差的核心问题，提出了一种名为ICLAD的全新范式。该框架利用音频语言模型（ALM）的上下文学习能力，实现了无需训练的快速适应。其核心是创新的**成对比较推理**策略：在离线阶段，引导ALM为每个样本同时生成“真实”和“伪造”的
每日研究速递
语音/音乐/音频论文速递 2026-04-21
共分析 34 篇语音/AI 论文