Max Bright
· LinkedIn
· Resume
· GitHub

Max Apollo Bright

I'm a Research Scientist at Meta, and I build Machine Learning models!

I work on the classifiers on Facebook and Instagram that detect violations of Meta's suicide and self-injury policies.

I specialize in computer vision and scientific computing. When I was a Research Engineer at Applied Research in Acoustics, I worked on few-shot learning, transformer pre-training, and active-learning methods for object detection and semantic segmentation. I finished my PhD in Electrical Engineering in 2022 at the University of Michigan, where I used wavelet transforms to speed up algorithms for simulating electromagnetic wave propagation. My advisor was Professor Eric Michielssen, and my dissertation is available here.

I also have a ton of hobbies! My current interests include:

  • Photography: I mainly focus on car photography, and I've also done portraits and fashion photoshoots.
  • Electronic Music Production: I make music heavily inspired by the trance, big-beat, and jungle soundtracks of late 90s/early 00s racing games.
  • Crafting: I enjoy sewing and knitting; I'm also experimenting with crochet and embroidery.
  • Car Racing: I frequently attend track days and autocross events.
  • Street Dancing: My style is a blend of house dance and Detroit jit.

Electronics and Software Projects

These are some of the projects I worked on in my free time when I was a student:
  • Synthesizer
  • I designed an analog synthesizer in LTSpice, and I prototyped hardware for a voltage-controlled oscillator and a voltage-controlled filter.

    VCO Printed Circuit Board
    VCO prototype

    VCO Prototype
    VCO prototype

    VCF Prototype


  • Real Time Convolution Plugin
  • Audio plugin for music production, made with the IPlug extension of the wdl-ol library. The plugin takes two real-time input tracks and convolves the most recent samples from both tracks together. It can be used for an effect similar to what a vocoder would produce.

    GitHub Repository
    Demonstration

  • MIDI Controller
  • Originally a simple circuit bending project to add distortion to a Casio SA-2 keyboard. I wired the keyboard to an Arduino, and was able to interpret the electrical signals as MIDI notes that I could use to control software synthesizers.

    Controlling a virtual instrument with the MIDI controller


  • Miscellaneous projects can be found on my GitHub account

Publications and Conference Presentations

  • J. Christian, M. Bright, J. E. Summers, A. Olson, T. C. Havens, “Generative EO/IR multi-scale vision transformer for improved object detection,” Proc. SPIE 13035, Synthetic Data for Artificial Intelligence and Machine Learning: Tools, Techniques, and Applications II, 130350N
  • M. Bright, J. Christian, R. Wood, J. E. Summers, “Unsupervised Vision-Transformer Training for Semantic Segmentation of Sparsely Labeled Datasets,” in Naval Applications of Machine Learning, San Diego, CA, 2024.
  • M. Bright, E. Michielssen, “Gabor Frame-Based Sparsification of 3D Long-Range Propagation over Terrain,” in IEEE International Symposium on Antennas and Propagation and USNC-URSI Radio Science Meeting, Denver, CO, 2022.
  • M. Bright, E. Michielssen, “Long-Range Propagation in 3D with Gabor Frame-Based Sparsification and Radiation Boundary Conditions,” in IEEE International Symposium on Antennas and Propagation and USNC-URSI Radio Science Meeting, Marina Bay Sands, Singapore, 2021.
  • M. Bright, E. Michielssen, “Time-Frequency Jigsaw Puzzle for Gabor Frame-Based Propagation over Long Ranges,” in IEEE International Symposium on Antennas and Propagation and USNC-URSI Radio Science Meeting, Marina Bay Sands, Singapore, 2021.
  • M. Bright, E. Michielssen, J. Kusuma, ‘Hybrid Split-Step Fourier – Finite Difference Parabolic Equation Solver for Modeling Propagation Over Complex Terrain’ in IEEE International Symposium on Antennas and Propagation and USNC-URSI Radio Science Meeting, Montreal, QC, 2020.
  • M. Bright, E. Michielssen, J. Kusuma, ‘Gabor Frame-Based Sparsification and Radiation Boundary Conditions for Parabolic Wave Equations’ in IEEE International Symposium on Antennas and Propagation and USNC-URSI Radio Science Meeting, Atlanta, GA, 2019.
  • M. Bright, A. Ishimaru, Y. Kuga, ‘Scattering from a Distribution of Rough Plates,’ Progress In Electromagnetics Research C, Vol. 95, 1-13, 2019.
  • M. Bright, A. Ishimaru, Y. Kuga, ‘Scattering from a Distribution of Rough Plates’ in USNC-URSI National Radio Science Meeting, Boulder, CO, 2018.
  • B. Walborn, Y. Liu, M. Bright, A. Ishimaru, Y. Kuga, ‘Modeling EM Wave Scattering from Tree Branches and Leaves’ in USNC-URSI National Radio Science Meeting, Boulder, CO, 2018.