Portfolio and research
Niterói, Brazil
Marcos Lázaro Romão Silveira
Electrical Engineering at UFF, with emphasis on electronics and machine learning
Most of what I do happens where a model meets the physical world. At LACOP, the optical communications laboratory at UFF, I apply learning methods to images and signals, and spend a good share of my time making them small enough to run on inexpensive hardware.
Along the way I have built a pedestrian detector for an ESP32 camera, gesture recognition that needs nothing beyond a webcam, and a communication app for children with cerebral palsy. What ties these together is a preference for systems someone can actually use, and for writing honestly about how they came to be.
- Now
- Undergraduate researcher at LACOP, UFF
- Focus
- Machine learning, computer vision, embedded ML
- Speaks
- Portuguese, English and Spanish
Projects
Read live from GitHub, so this index moves whenever the code does.
Research
The work at LACOP and the questions behind it.
My work at LACOP asks how statistical learning can make sense of the data physical systems produce, from camera frames to optical and electrical signals. Recently that has meant vision models compact enough for a microcontroller's memory, alongside applied projects in which AI supports people with motor or speech impairments.
Lines of inquiry
- Computer vision on embedded hardware
- TinyML, quantization and model compression
- Machine learning for optical and electrical signals
- Assistive technology and augmentative communication
Publications
Writing
A place for essays and lab notes.
Contact
Open to research collaborations, internships and thoughtful questions.
If you are working on something where machine learning has to leave the notebook, run on real hardware and serve the people who use it, I would like to hear about it.
Email is the most reliable way to reach me, and I usually reply within a couple of days. The other channels are listed below.