CV

Research interests

• Automatic Speech Recognition:
  • Noise-robust
  • Adaptation
  • Multi-speaker
  • Multi-lingual
• Speech Enhancement:
  • Front-end for ASR
  • Systems ensemble
  • Probabilistic (generative) model
  • Multi-model
• Speech Separation:
  • Target speaker extraction
  • Blind source separation

Education

• 2021 - 2024, Ph.D. in Informatics, Kyoto University
  • Robust Automatic Speech Recognition
  • Supervisor: Tatsuya Kawahara
• 2018 - 2021, M.Phil. in Computer Science, Tianjin University
  • Speech Enhancement
  • Supervisor: Longbiao Wang
• 2014 - 2018, B.Sc. in Computer Science, Southwest Jiaotong University
  • Supervisor: Zhixiong Di

Work experience

• 2024.10 - Present: Researcher
  • Kyoto University
  • Speech signal processing
• 2024.04 - 2024.09: Research Fellow
  • Kyoto University
  • Speech signal processing
• 2023.08 - 2023.09: Research Intern
  • NTT
  • Speech signal processing
  • Mentor: Naoyuki Kamo, Shoko Araki, Tomohiro Nakatani, Marc Delcroix
• 2023.01 - 2023.03: Research Intern
  • Sony
  • Speech signal processing
  • Mentor: Kazuki Shimada, Shusuke Takahashi
• 2021.08 - 2022.01: Research Assistant
  • Tianjin University
  • Speech signal processing

Honors

• 2022.04 - 2024.03: Fellowship
  • awarded by Japan Science and Technology Agency (JST)

Language skill

• Chinese (native)
• English (fluent)

Publications

Adapting Pretrained Speech Recognition Models for Code-Switching through Encoding Refining and Language-Aware Attention-based Decoding

Reducing the Gap between Pretrained Speech Enhancement and Recognition Models Using a Real Speech-Trained Bridging Module

Serialized Speech Information Guidance with Overlapped Encoding Separation for Multi-Speaker Automatic Speech Recognition

Error Correction by Paying Attention to Both Acoustic and Confidence References for Automatic Speech Recognition

Dual-path Adaptation of Pretrained Feature Extraction Module for Robust Automatic Speech Recognition

Speech Emotion Recognition with Multi-level Acoustic and Semantic Information Extraction and Interaction

Waveform-domain Speech Enhancement Using Spectrogram Encoding for Robust Speech Recognition

Ensemble Inference for Diffusion Model-based Speech Enhancement

Diffusion-Based Speech Enhancement with Joint Generative and Predictive Decoders

Enhancing Two-stage Finetuning for Speech Emotion Recognition Using Adapters

Investigation of Adapter for Automatic Speech Recognition in Noisy Environment

Extending Audio Masked Autoencoders Toward Audio Restoration

Time-domain Speech Enhancement Assisted by Multi-resolution Frequency Encoder and Decoder

Adaptive Attention Network with Domain Adversarial Training for Multi-Accent Speech Recognition

Subband-Based Spectrogram Fusion for Speech Enhancement by Combining Mapping and Masking Approaches

Fusing Multiple Bandwidth Spectrograms for Improving Speech Enhancement

Self-Distillation Based on High-level Information Supervision for Compressing End-to-End ASR Model

Language-specific Characteristic Assistance for Code-switching Speech Recognition

Monaural speech enhancement based on spectrogram decomposition for convolutional neural network-sensitive feature extraction

Spectrograms Fusion-based End-to-end Robust Automatic Speech Recognition

Simultaneous Progressive Filtering-based Monaural Speech Enhancement

Speech Dereverberation Based on Scale-aware Mean Square Error Loss

Singing Voice Extraction with Attention-Based Spectrograms Fusion

Spectrograms Fusion with Minimum Difference Masks Estimation for Monaural Speech Dereverberation

Environment-Dependent Attention-Driven Recurrent Convolutional Neural Network for Robust Speech Enhancement

Reviewer

• Jounral
  • IEEE/ACM Trans. ASLP
  • Speech Communication
• Conferences or Workshops
  • IEEE-ICASSP
  • INTERSPEECH
  • APSIPA ASC
  • IEEE-SLT
  • WASPAA