Kaifeng Lu

Location: Berkeley, CA

Location: University of California, Berkeley

klu15 at berkeley dot edu

Hi! I an undergraduate student in Mathematics and minor in D.S. who currently studies at UC Berkeley. Previously, I was a student at the University of Rochester in Math+CS.

I am currently working as an undergraduate research apprientices at professor Lyu’s lab. I have broad interests in medical imaging and deep learning, especially on medical image processing and LSTM.

You can find my full CV here: html | pdf

I am actively looking for undergraduate research assistant position either remote or in-person. Please email me if you are interested.

Visitor’s Map

✈️News

Sep 21, 2025	🔥 A demo that I contribute has been submitted to ISMIR2025!

✍️Latest posts

Jun 11, 2025	I just create my first personal website!
May 15, 2025	转学log与心得分享

selected publications

MusicSem: A Dataset of Music Descriptions on Reddit Capturing Musical Semantics

Rebecca Salganik, Teng Tu, Fei-Yueh Chen, and 8 more authors

2025

Abs Bib HTML PDF Code

MusicSem is a 32k-pair language–audio dataset mined from organic Reddit discussions, designed to capture how listeners naturally describe music. The captions are organized into five semantic categories—descriptive, atmospheric, situational, metadata-related, and contextual—so models can be trained and evaluated for sensitivity to nuanced, human-centered intent in text-to-music tasks and cross-modal retrieval.

@demo{salganik2025musicsem,
  title = {MusicSem: A Dataset of Music Descriptions on Reddit Capturing Musical Semantics},
  author = {Salganik, Rebecca and Tu, Teng and Chen, Fei-Yueh and Liu, Xiaohao and Lu, Kaifeng and Luvisia, Ethan and Duan, Zhiyao and Salha-Galvan, Guillaume and Kahng, Anson and Ma, Yunshan and Kang, Jian},
  booktitle = {Demo, 26th International Society for Music Information Retrieval Conference},
  year = {2025},
  journal = {https://ismir2025program.ismir.net/lbd_388.html},
  dataset = {https://huggingface.co/datasets/Rsalga/MusicSem},
}