Apply
We welcome students who are passionate about audio, speech processing, machine learning, and generative AI research.
Application Info
We welcome students who are passionate about audio, speech processing, machine learning, and generative AI research. Applicants are encouraged to send: • CV / Resume • Research interests • Project or coding experience • Relevant coursework or activity records Links to GitHub repositories, portfolio pages, papers, demos, or other technical materials are highly encouraged. We value curiosity, consistency, and a strong motivation to learn and build. For inquiries or applications, please contact: jwoo@kaist.ac.kr
Students interested in joining the lab are encouraged to explore general research foundations in audio and speech AI.
Study
Students interested in joining the lab are encouraged
to explore research topics related to:
• Speech & Audio
speech enhancement
sound source separation
target sound extraction
speech recognition
• Spatial & Acoustic Audio Systems
spatial audio
room acoustics & RIR
microphone array signal processing
• Deep Learning for Audio
PyTorch / PytorchLightning
CNN / RNN / Transformers / Mamba
self-supervised representation learning
• Generative Audio Modeling
diffusion models
flow matching / meanflow
audio generation & restoration
• Audio Foundation Models
WavLM / HuBERT / Whisper
multimodal audio-language models
These topics are intended as preliminary areas of exploration for students interested in audio and speech AI research.
Application
Students are encouraged to engage in research projects related to audio and speech technologies.
• Sound Event Localization & Detection
Rebuilding Class / DoA decoders with Transformer-based architectures
Applying multi-scale structures to DeFT-based blocks (SEMamba++)
Exploring magnitude / phase mu-law compression & decompression techniques
• Multichannel Target Sound Extraction with Spatial queries
Simulating complex multi-source acoustic environments
Designing spatial queries using target direction and distance information
Developing discriminative and generative audio extraction models
Investigating robust spatial representation learning for multichannel audio
• Spatial Audio Understanding for Large Audio Language Models
Building spatial audio understanding benchmarks for multimodal LLMs
Generating simulated spatial audio datasets using audio simulation tools
Exploring multimodal learning with spatial acoustic information
• Spatial Audio Rendering
Simulating Room Impulse Responses in complex rooms
Developing models for Room Impulse Response generation
Exploring evaluation metrics related to human auditory perception
Students may gain hands-on experience with research tasks and projects through participation in audio and speech AI research.