graph LR
ML equivalents column
subgraph ML_Equivalents
conv[Convolutional layers\nlocal filters, shared weights]
stft[STFT / Filterbanks\nspectral decomposition]
patch[Patch embeddings / Local kernels\nlearned local features]
norm[Divisive normalization / BatchNorm\ngain control, contrast normalization]
deep[Deep CNNs / Transformers\nhierarchical feature extraction, attention]
sparse[Sparse coding / Autoencoders / L1\nhigh-dimensional sparse reps]
end
Cross-links (additional analogies)
R --> norm
conv --> deep
stft --> patch
norm --> deep
sparse --> deep
classDef bio fill:#eef6ff,stroke:#0b66ff,color:#012a4a;
classDef ml fill:#fff7e6,stroke:#ff8c00,color:#663300;
class R,C,S,E,H,O bio;
class conv,stft,patch,norm,deep,sparse ml;
%% compact notes
linkStyle 0 stroke:#888,stroke-width:1.5px
linkStyle 1 stroke:#888,stroke-width:1.5px
linkStyle 2 stroke:#888,stroke-width:1.5px
linkStyle 3 stroke:#888,stroke-width:1.5px
linkStyle 4 stroke:#888,stroke-width:1.5px
linkStyle 5 stroke:#888,stroke-width:1.5px