cutmix does: replaces a patch of one image with a patch from another

label smoothing does: replaces one-hot [0,0,1,0] with [0.025, 0.025, 0.925, 0.025]

How can a computer learn without being told exactly what to do?

operator fusion does at the compiler level: merges adjacent ops to reduce memory traffic

Did you know your computer can write faster code than you?

convolution (f * g)(t) = ∫f(τ)g(t-τ)dτ

How do different shapes combine to create new patterns?

Fréchet inception distance

Fréchet inception distance (FID) compares image distributions

mixup does: trains on convex combinations of pairs: x̃=λx_i+(1-λ)x_j

Ever wondered how to blend two colors perfectly?

fused kernels do

Why wait for a computer to finish one task before starting another?