Sid Banerjee

I'm a senior at UC Berkeley studying Electrical Engineering & Computer Science. My academic interests are in digital signal processing, acoustics/computational musicology (specifically MIR), and machine learning.

My work is mainly focused on DSP signal chains and deep learning algorithms to better understand, characterize, and create music. I'm aim not to develop end-to-end generative AI models but rather use AI/ML as a means of "crunching musical data" to reveal interesting, hidden structures & power algorithms in human-centric composition tools.

I create music/art via Max (+ TouchDesigner) and Ableton. My Max patches are usually techniques for synthesis, including cross synthesis, subtractive synthesis, and granular synthesis. I prefer TouchDesigner for audio-reactive visuals (over Jitter) and use OSC to connect to my Max patches. The music I produce with Ableton can be best described as a blend of electronic/hyperpop, cloud rap, and jazz.

At university, I conduct research @ CNMAT and BAIR on the following projects:

Studying the impact of deep learning modules on generated compositions to understand how context and memory benefit/impair creativity.
Morphological and perceptually driven Matching Pursuit for computer-assisted orchestration.
Conversational behavior modeling through predictive transformer models and dynamic Graph of Thought generation

I'm also a recitation/discussion leader on course staff of ELENG66 (Signal Processing & Applied Linear Algebra); I work with the team to write and teach course content.

Talk [jazz, art, computer science, math, etc.] with me: sidbanerjee[at]berkeley.edu.

Sid Banerjee

Musician and Engineer, in the Media Arts & Technology

Student @ UC Berkeley, CO 2027