Research Assistant – NYU MARL Music & Social Change Lab
Research Collaborator – RAVE Experimental Music Project
Research Intern – Tencent Music Entertainment Group
Contributing to Dr. Carlos Chirinos’s paper Leveraging AI and Data Mining for Automated Retrieval of Prosocial Songs’ Lyrics by designing a GPT-based augmentation pipeline and developing a Cross-Attention Fusion Model for automated topical detection with a focus on interpretability and bias analysis in music-AI systems.
Collaborating with Prof. Matthew Ostrowski on optimizing IRCAM’s RAVE neural synthesis models and designing data pipelines for 1920s–30s spoken-word corpora to achieve expressive timbre-matching and latent-space control in the Max/MSP nn~ framework.
Supported the AI Tone Magician team by evaluating personalized singing-voice conversion models, organizing user studies on vocal-identity retention, and contributing to real-time karaoke feature testing to benchmark model latency and lip-sync alignment across deployment settings within the WeSing ecosystem.
Research Intern – TROY Information
Developed multimodal verification models integrating speech, facial, and behavioral embeddings via temporal-attention alignment for interpretable cross-modal analysis.
Research Intern – Tencent Games
Supported reinforcement-learning experiments for in-game AI behavior research and analyzed engine performance and player–AI interaction logs to optimize real-time gameplay latency, synchronization, and responsiveness.
Research Intern – OriginMind Technology
Implemented transformer-based symbolic-music generation conditioned on text prompts for melody, rhythm, and texture control.