Matthew Lam
I'm a 2nd-year Engineering Science student at the University of Toronto. I'm interested in software and hardware for visual computing, sensing and computational imaging. I'm also interested in sustainability and clean energy. In addition, I'm a member of the University of Toronto Aerospace Team (Space Systems), on the Electrical subteam (and previously in RF Communications).
Projects
Modified 3D Gaussian Splatting pipeline to infer albedo and surface normals, allowing for rendering views under novel lighting conditions by applying a Lambertian reflectance model. Plus a custom dataset.
Custom-designed U-Net for motion deblurring, trained on the GoPro dataset. I also ran an ablation comparing NAF blocks with SimpleGate activations to modified blocks with GELU activations.
This was my first computer vision project as well as my first ML project. It consists of a convolutional neural network that classifies images of clothing items. It is trained on the Fashion-MNIST dataset.
Matboard beam bridge spanning 1.2 m and withstanding a 580 N load. We wrote a Python script to calculate factors of safety of our design, and predict the bridge's performance.
I designed and ran a water electrolysis experiment to determine the effect of varying NaOH concentration on rate of water electrolysis. I measured rate using electrical current and rate of H2 production.
Comparison of nearest-neighbor, bilinear, bicubic and bicubic spline interpolation for image resizing. I applied these methods to an image in MATLAB, then compared results visually and with mean squared error.
Exploration of models of musical consonance based on beating and roughness, including William Sethares's model (1999). Illustration of the model using FFT in MATLAB to compute frequency spectra.
Artwork incorporating visual and musical components using graphs of mathematical equations. Placed top 20 in the 2022 Desmos Global Math Art Contest (age 15-16 category).
Miscellaneous
- Justification for Method of Comparing Coefficients: Proof that justifies the use of the method of comparing coefficients (e.g. \(ax^2+bx+c=1x^2+2x+3\) implies \(a=1, b=2, c=3\)).
- Useful Equations for ESC195 Term Test #2: Compilation of important but easily-forgettable equations relevant to the second term test of ESC195.
- Matrix Multiplication and Gaussian Elimination Review: Summary of useful facts about matrix multiplication, and a brief summary of the method of Gaussian Elimination.