Current Projects
-
Differentiable Auditory Processing
new
Implemented a differentiable version of auditory processing model based on signal processing.
-
Rhythm and Language Discrimination
Humans can discriminate between certain languages right after birth. Speech models suggest rhythm may be irrelevant to this.
-
Robust Model Comparison
Developed a method using U-statistics and delta method to derive unbiased estimate for model comparison.
(Click on project for related publications.)
About me
I am interested in simulating how people perceive and learn about the world. Within this frame, I have been studying auditory perception and processing as a particular medium, and examining how people learn, use, and represent speech in their brain.
Currently, I am a PhD student at the University of Maryland, with affiliations in NACS (Neuroscience and Cognitive Science), the Department of Computer Science, and the Department of Linguistics. My advisors are Drs. Naomi Feldman and Ramani Duraiswami. I also pursued a MS degree in Computer Science concurrently with my PhD in Neuroscience and Cognitive Science.
Aside from research, I am a pianist who plays what I love like a broken record. Other than performing, I enjoy studying theories of music and music history, occasionally composing in the old style. I take music not as a career but very seriously, because as Seneca says:
Mihi crede, verum gaudium res severa est. (Sen. Ep. 23.4)
Chronologically, I am from Kunming, a city in Southwest China. Intriguingly, the main dialect of Kunming is a Northern dialect, with close relation to standard Mandarin. Here is a little project on my hometown dialect. I graduated from the University of Rochester in 2020, with a degree in Brain & Cognitive Sciences (B.S.) and another in Linguistics (B.A.).
Other/Past Projects
-
Input to Phonetic Learning Models
Number and composition of speakers affect the outcome of a model of human phonetic learning.
-
Foreign-Accented Speech Corpus
A corpus containing native and Mandarin-accented English, including isolated words and connected speech.
-
Evolution of Musical Cadences
Semester project simulating the change of cadences in music, from the 1400s to the Renaissance.
-
Confirmation Bias and Information Structure
Replication and expansion on an existing model of confirmation bias, with human data.
Contact me
Credits
This website uses jekyll, with theme modified from Nice blog by Gabriel Chen. My customizations are also inspired by themes Moonwalk and Researcher.