Research
Books Link to heading
I became deeply interested in biological neural networks during my time at Stanford. I stumbled upon active research on the topic being performed by Dr. Bernard Widrow during his course on adaptive signal processing. For three years after the course, I partnered with him to develop neural networks with local learning rules inspired by homeostatic regulation of firing rate. The algorithms extended to other homeostatic systems such as thermoregulation, blood pressure, and even plant growth. Dr. Widrow came up with his ideas and we discussed them in great depth at his residence or the faculty club. I assisted him by running computer simulations, performing literature reviews, and ultimately transcribing his hand-written words into a book published by Springer.
Papers and reports Link to heading
% LMS: A stochastic gradient algorithm inspired by neurobiology
2019 | Stanford CS229 | paper link | poster link
Audio bandwidth extension using a normalizing flow model (WaveGlow)
2020 | Stanford CS236 | paper link
GRU Voice Activity Detector on ARM Cortex M4 using Tensorflow
2019 | Stanford CS230 | paper link | poster link
Speech denoising using adaptive filtering
2019 | Stanford EE373 | github link
Pyramid vector quantization for audio compression
2020 | Stanford EE376C | paper link | slidedeck link
Mitigating catastrophic interference in multi-task learning
2020 | Stanford CS376C | paper link
Audio-based gesture detection on the Neosensory Buzz wristband (Cortex M4)
2019 | Stanford EE292D | paper link | video demo | slidedeck link
Tactile representations of image signals
2018 | Stanford EE376A | paper link
Wireless performance characterization of fabric antennas
2013 | Rice CMC Lab | paper link
Patents Link to heading
US20200209975A1: Method and system for providing adjunct sensory information to a user
US11614802B2: Method and system for haptic stimulation
Talks Link to heading
Python in MS Excel
2014 | PyTexas | video link
How understanding the human brain can impact our lives
2015 | TEDxYouth@DPSMIS | video link