CV

Research Intern

Dolby Laboratories, Inc.

Video to Spatial Audio (FOA) Generation via Latent Diffusion

• Spatial Audio
• Multi-modal
• Audio Generation
• Diffusion

Acoustic Mapping Intern

Dolby Laboratories, Inc.

Robust User Localization in Acoustic Mapping via Speech Enhancement

• Spatial Audio
• Localization

Doctor of Philosophy

New York University

Computer Science

• Machine Learning
• Computer Vision
• Music Information Retrieval
• Information Visualization
• 3D Audio
• Digital Signal Processing

Master of Science

Columbia University

Data Science

• Machine Learning
• Deep Learning and Neural Networks
• Probability and Statistics
• Reinforcement Learning
• Statistical Inference and Modeling
• Algorithms for Data Science
• Computer Systems
• Sonic and Visual Representations of Data
• Sound: Advanced

Bachelor of Engineering

Wuhan University

Energy Engineering (Thermodynamics)

• Maths: Advanced Mathematics, Linear Algebra, Probability
• Mechanics: Theoretical Mechanics, Mechanical Design, Material Mechanics, Fluid Mechanics, Quantum Mechanics
• Electronics: Electrical Engineering and Electronics Techniques, Principle of Automatic Control
• Dynamics: Thermodynamics, Heat Transfer, Multi-scale Modeling and Simulation
• Programming: C, C++, Java

Spatial Scaper: A Library to Simulate and Augment Soundscapes for Sound Event Localization and Detection in Realistic Rooms

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

We present SpatialScaper, a library for SELD data simulation and augmentation.

SAMO: Speaker Attractor Multi-Center One-Class Learning For Voice Anti-Spoofing

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

In this work, we propose speaker attractor multi-center one-class learning (SAMO), which clusters bona fide speech around a number of speaker attractors and pushes away spoofing attacks from all the attractors in a high-dimensional embedding space.

SoniZen

Mindful Meditation Experience via Data Sonification

• Designed a multi-modal real-time plug-in with OSC and MAX for Live with gestural, visual, and auditory input as sensor signals
• Mapped signals from wearables (watches and earphones) to controlled parameters with neural networks for live music performance

Voice Anti-Spoofing and Audio Deepfake Detection

2022 Summer Internship @AIR, UoR

• Designed a novel loss function with an algorithm for speaker attractor multi-center one-class supervised learning with 120K voice data
• Refined generalizability of audio spoofing detection to achieve SOTA EER by 38% relative improvement
• Illustrated model behaviors in cluster representation learning and classification through ablation and UMAP and t-SNE embedding visualization
• Leveraged cyclic learning rate and hyper-parameter tuning techniques to improve convergence of training process

Embodied Multi-Modal Machine Listening in Audio-Visual Navigation

2022 Summer Internship @MARL, NYU

• Instrument audio/image CNN and Transformer with reinforcement learning on HPC clusters in semantic audio-visual navigation
• Developed an API for audio feature extraction for baseline models to perform transfer learning in holistic downstream evaluation

Soundscape Simulation, Augmentation and Visualization

A Python Library for Soundscape Generation

• Developed a Python library for data simulation, augmentation, spatialization, and visualization of spatial audio
• Conducted ablation studies with a DCASE SELD challenge model to manifest 37% improvement of augmentation over baseline

Acoustic Spatial Visualizer

A Python Library for Soundscape Visualization

• Generate spatial audio with SpatialScaper and visualize its moving track with DeepWave
• The visualizer uses APGD algorithm to takes in a 32-channel spatialized audio and outputs a 2D/3D energy map

Neuro-Harmonilizer

A Python Library for Soundscape Visualization

• It maps any chord to a polar coordinates ϕ, ρ, where ϕ means the color orientation and the ρ means the tension class within the total 31 classes.

Merit Student Scholarship

Wuhan University

Hualin Machinery First-Class Scholarship

Wuhan University

Relabo Innovation Scholarship

Wuhan University

Outstanding Student

Wuhan University