Shinji Watanabe
Associate Professor
Studies speech recognition, speech enhancement and spoken language processing, including the ESPnet toolkit.
Shinji Watanabe's lab at CMU works on speech recognition & spoken language, speech enhancement & signal processing and recognition, detection & segmentation. 267 papers since 2023, 13 at top venues, mostly TASLP and Findings of EMNLP. Applications close .
At a glance
- papers, 2023–2026
- 70/ year
- at T1 venues
- 5%
- PhD students
- —
- MS students
- —
- alumni listed
- —
Papers by yearfrom the lab homepage, 2023–2026
- 59 papers2026 · T1 1 · collecting
- 68 papers2025 · T1 5
- 66 papers2024 · T1 2
- 74 papers2023 · T1 5
T1 venuesOtherVenues: TASLP, Findings of EMNLP, Findings of ACL, TMLR
Research focuspaper titles that use each topic's words
- Speech recognition & spoken language202 papers
- Speech enhancement & signal processing156 papers
- speech153 papers
- Recognition, detection & segmentation51 papers
- speech enhancement17 papers
Paper list
From the lab homepage.
202659 papers
- Proc. Interspeech'26Adapting Text LLMs to Speech via Multimodal Depth Up-Scaling
- Proc. ICML'26ADEPT: RL-Aligned Agentic Decoding of Emotion via Evidence Probing Tools — From Consensus Learning to Ambiguity-Driven Emotion Reasoning
- Proc. Interspeech'26An Efficient vLLM-Based Inference Pipeline for Unified Audio Understanding and Generation
- Proc. Interspeech'26An Empirical Recipe for Universal Phone Recognition
- Computer Speech & Language, p. 101813An End-to-End Integration of Speech Separation and Recognition with Self-Supervised Learning Representation
- Proc. Interspeech'26ANCHOR: Autoregressive Non-intrusive Chunk-Ordered Refinement for Joint Multi-Resolution Speech Quality Modeling
- Proc. SLT'26ATLAS: Agentic ASR Tuning for Research Labor Scaling
- Proc. ICML'26AudioChat: Unified Audio Storytelling, Editing, and Understanding with Transfusion Forcing
- Proc. Interspeech'26Bagpiper-Edit: Zero-Shot Open-Ended Audio Editing via Rich-Caption
- Proc. Interspeech'26Bagpiper-TTS: Natural Language Guided Universal Speech Synthesis
- Proc. COLM'26Bagpiper: Solving Open-Ended Audio Tasks via Rich Captions
- Findings of EMNLPBenchmarking Speech-to-Speech Translation Models
202568 papers
- Proc. ICASSP'262025 URGENT Speech Enhancement Challenge Multilingual P.808 Listening Tests: Approach And Results
- TASLP (T1)Aligning Speech to Languages to Enhance Code-switching Speech Recognition
- Proc. ISMIR'25Aligning Text-to-Music Evaluation with Human Preferences
- Proc. NeurIPS'25ARECHO: Autoregressive Evaluation via Chain-Based Hypothesis Optimization for Speech Multi-Metric Estimation
- Proc. ASRU'25 DemoAURA: Agent for Understanding, Reasoning, and Automated Tool Use in Voice-Driven Tasks
- Proc. ICASSP'25Bridging Speech and Text Foundation Models with ReShape Attention
- Proc. Interspeech'25Chain-of-Thought Training for Open E2E Spoken Dialogue Systems
- Proc. ICLR'25Context-aware Dynamic Pruning for Speech Foundation Models
- Proc. Interspeech'25Context-Driven Dynamic Pruning for Large Speech Foundation Models
- Proc. ASRU'25Continual Pre-training for Codec-Based Speech LLMs: Balancing Understanding and Generation
- Proc. Interspeech'25CS-FLEURS: A Massively Multilingual and Code-Switched Speech Dataset
- Proc. Interspeech'25DiceHuBERT: Distilling HuBERT with a Self-Supervised Learning Objective
202466 papers
- Proc. ICASSP'24AUGSUMM: TOWARDS GENERALIZABLE SPEECH SUMMARIZATION USING SYNTHETIC LABELS FROM LARGE LANGUAGE MODELS
- Proc. Interspeech'24Beyond Performance Plateaus: A Comprehensive Study on Scalability in Speech Enhancement
- Proc. ICASSP'24BOOSTING UNKNOWN-NUMBER SPEAKER SEPARATION WITH TRANSFORMER DECODER-BASED ATTRACTOR
- Proc. Interspeech'24Can you Remove the Downstream Model for Speaker Recognition with Self-Supervised Speech Features?
- Proc. SLT'24Codec-SUPERB @ SLT 2024: A lightweight benchmark for neural audio codec models
- Proc. ICASSP'24CONTEXTUALIZED AUTOMATIC SPEECH RECOGNITION WITH ATTENTION-BASED BIAS PHRASE BOOSTED BEAM SEARCH
- Proc. SLT'24Contextualized Automatic Speech Recognition with Dynamic Vocabulary
- Proc. Interspeech'24Contextualized End-to-End Automatic Speech Recognition with Intermediate Biasing Loss
- Proc. Interspeech'24Convolution-Augmented Parameter-Efficient Fine-Tuning for Speech Recognition
- Proc. ICASSP'24CROSS-MODAL MULTI-TASKING FOR SPEECH-TO-TEXT TRANSLATION VIA HARD PARAMETER SHARING
- Proc. IJCAI'24Cross-Talk Reduction
- Proc. Interspeech'24Decoder-only Architecture for Streaming End-to-end Speech Recognition
202374 papers
- Proc. Interspeech'234D ASR: Joint modeling of CTC, Attention, Transducer, and Mask-Predict decoders
- Proc. Interspeech'23A Comparative Study on E-Branchformer vs Conformer in Speech Recognition, Translation, and Understanding Tasks
- Computer Speech & LanguageA Dilemma of Ground Truth in Noisy Speech Separation and an Approach to Lessen the Impact of Imperfect Training Data
- Proc. Interspeech'23A New Benchmark of Aphasia Speech Recognition and Detection Based on E-Branchformer and Multi-task Learning
- Proc. ASRU'23A Single Speech Enhancement Model Unifying Dereverberation, Denoising, Speaker Counting, Separation, and Extraction
- Proc. ICASSP'23A Unified One-Shot Prosody and Speaker Conversion System with Self-Supervised Discrete Speech Units
- Proc. AAAI'23A Vector Quantized Approach for Text to Speech Synthesis on Real-World Spontaneous Speech
- Proc. ICASSP'23Align, Write, Re-order: Explainable End-to-End Speech Translation via Operation Sequence Generation
- Proc. ICASSP'23Articulatory Representation Learning Via Joint Factor Analysis and Neural Matrix Factorization
- Proc. ICASSP'23Avoid Overthinking in Self-Supervised Models for Speech Recognition
- Proc. Interspeech'23BASS: Block-wise Adaptation for Speech Summarization
- Proc. ICLR'23BAYES RISK CTC: CONTROLLABLE CTC ALIGNMENT IN SEQUENCE-TO-SEQUENCE TASKS
Where alumni went
Alumni are not listed on the lab homepage.
Apply to CMUFall 2027
CMU SCS Computer Science PhD
- Round
- Single round
- Opens
- GRE
- See program page
- English
- TOEFL / IELTS / Duolingo (F-1/J-1 non-native speakers)
- Fee
- USD 100 (USD 80 by early deadline)
Confirmed on the program page.
Similar labsnearest by research summary
- GBEdinburgh
Peter Bell
- Speech recognition & spoken language
- USGeorgia Tech
Biing Hwang Juang
- Speech recognition & spoken language
- CNPeking
Xihong Wu
- Speech recognition & spoken language
- CAToronto
Gerald Penn
- Multilingual, translation & linguistics
- CNSJTU
Xie Chen
- Speech recognition & spoken language
- CNTsinghua
Liang He
- Speech recognition & spoken language
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The member list could not be found on the lab site.
- The alumni list could not be found on the lab site.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Not collected
- Alumni
- Not collected
Tiers are venue classifications, not lab rankings.
CMUcloses