Interspeech 2025

1,156 papers

Top Research Directions

Research directions most represented by Interspeech 2025 papers.

Partitioning Variance with Linear Mixed Effects Models 44
Learning High-Level Acoustic Structures from Unlabeled Audio 42
Mitigating the Acoustic Mismatch in Atypical Speech Recognition 35
Fine Tuning Speech Models for Specialized Domains 28
Factorizing Speech into Content Timbre and Prosody 25
Overcoming Data Scarcity in Low-Resource Speech Recognition 25
Overcoming the Modality Gap in Audio Language Models 19
Synergistic Multi Task Learning for Speech and Audio 18
The Perceptual Evaluation Gap in Generative Audio 18
Temporal Alignment of Speech and Text 17
Fine-Grained Prosodic Control for Expressive Speech Synthesis 17
Generalizing from Limited Cohorts in Speech Research 17
Speech Analysis for Early Detection of Cognitive Decline 16
Zero Shot Speaker Generalization for Speech Synthesis 16
Mitigating Data Scarcity in Affective Speech Models 16

Top Authors

AuthorPapers
Shinji Watanabe22
Haizhou Li10
Shrikanth Narayanan10
Yanmin Qian10
Dong Yu9
Hung-yi Lee8
EngSiong Chng7
Sabato Marco Siniscalchi6
Xiaofen Xing6
Xixin Wu6
Bin Ma5
Jagadeesh Balam5
Jian Luan5
Joon-Hyuk Chang5
Kyogu Lee5
Man-Wai Mak5
Najim Dehak5
Shujie HU5
Yifan Peng5
Chenxing Li4

Impactful Papers

Papers from Interspeech 2025 that shaped research clusters.

Foundational

Influential

Papers

Title
LightL2S: Ultra-Low Complexity Lip-to-Speech Synthesis for Multi-Speaker Scenarios
Lightweight and Robust Multi-Channel End-to-End Speech Recognition with Spherical Harmonic Transform
Lightweight Front-end Enhancement for Robust ASR via Frame Resampling and Sub-Band Pruning
Lightweight Speech Enhancement for Mandarin Esophageal Speech
Lightweight Speech Enhancement Model Based on Harmonic Attention and Phase Estimation with Skin-Attachable Accelerometer
LinearVC: Linear Transformations of Self-Supervised Features Through the Lens of Voice Conversion
Linguistic Masking and Its Release in Simulated Electric-acoustic Hearing
LiRI Corpus Platform: Demonstration of a Web-Based Infrastructure for Multimodal Corpus Analysis
Listen, Analyze, and Adapt to Learn New Attacks: An Exemplar-Free Class Incremental Learning Method for Audio Deepfake Source Tracing
LiSTEN: Learning Soft Token Embeddings for Neural Audio LLMs
Listen through the Sound: Generative Speech Restoration Leveraging Acoustic Context Representation
LIST: Language-Independent Speech Token for Multilingual Speech Synthesis with Language Models
LitMAS: A Lightweight and Generalized Multi-Modal Anti-Spoofing Framework for Biometric Security
LLM-based Generative Error Correction for Rare Words with Synthetic Data and Phonetic Context
LLM-based phoneme-to-grapheme for phoneme-based speech recognition
LLM-Synth4KWS: Scalable Automatic Generation and Synthesis of Confusable Data for Custom Keyword Spotting
Location-Aware Target Speaker Extraction for Hearing Aids
LombardTokenizer: Disentanglement and Control of Vocal Effort in a Neural Speech Codec
Long-Context Speech Synthesis with Context-Aware Memory
Loquacious Set: 25, 000 Hours of Transcribed and Diverse English Speech Recognition Data for Research and Commercial Use
Low Complex IIR Adaptive Hear-Through Ambient Filtering for Overcoming Practical Constraints in Earbuds
LRBA: Stealthy Backdoor Attacks on Speech Classification via Latent Rearrangement in VITS
LSCodec: Low-Bitrate and Speaker-Decoupled Discrete Speech Codec
LSPnet: an ultra-low bitrate hybrid neural codec
M3L: A Multi-Modal and Multi-Lingual Depression Detection Framework
MADUV: The 1st INTERSPEECH Mice Autism Detection via Ultrasound Vocalization Challenge
Mamba-based Hybrid Model for Speech Enhancement
MASV: Speaker Verification with Global and Local Context Mamba
MATER: Multi-level Acoustic and Textual Emotion Representation for Interpretable Speech Emotion Recognition
MDDM: A Multi-view Discriminative Enhanced Diffusion-based Model for Speech Enhancement
Medusa: A Multimodal Deep Fusion Multi-Stage Training Framework for Speech Emotion Recognition in Naturalistic Conditions
Mel-McNet: A Mel-Scale Framework for Online Multichannel Speech Enhancement
MelRe: Vision-Based Mel-Spectrogram Restoration
Meta-Learning Approaches for Speaker-Dependent Voice Fatigue Models
Meta-PerSER: Few-Shot Listener Personalized Speech Emotion Recognition via Meta-learning
MFLA: Monotonic Finite Look-ahead Attention for Streaming Speech Recognition
MIKU-PAL: An Automated and Standardized Multimodal Method for Speech Paralinguistic and Affect Labeling
Mimic Blocker: Self-Supervised Adversarial Training for Voice Conversion Defense with Pretrained Feature Extractors
Mispronunciation Detection Without L2 Pronunciation Dataset in Low-Resource Setting: A Case Study in Finland Swedish
MiSTR: Multi-Modal iEEG-to-Speech Synthesis with Transformer-Based Prosody Prediction and Neural Phase Reconstruction
Mitigating Audiovisual Mismatch in Visual-Guide Audio Captioning
Mitigating Language Mismatch in SSL-Based Speaker Anonymization
Mitigating Non-Target Speaker Bias in Guided Speaker Embedding
Mitigating Overfitting During Speech Foundation Model Fine-tuning: Applications to Dysarthric Speech Detection
Mitigating Subgroup Disparities in Multi-Label Speech Emotion Recognition: A Pseudo-Labeling and Unsupervised Learning Approach
Mixture of LoRA Experts for Low-Resourced Multi-Accent Automatic Speech Recognition
MMLoRA: Multitask Memory Parameter-Efficient Fine-Tuning for Multimodal SER
MM-MovieDubber: Towards Multi-Modal Learning for Multi-Modal Movie Dubbing
Modality-Agnostic Multimodal Emotion Recognition using a Contrastive Masked Autoencoder
Modality-Specific Speech Enhancement and Noise-Adaptive Fusion for Acoustic and Body-Conduction Microphone Framework

« prev Page 13/24 next »