GraFT (Graph-Filtered Temporal) is a signal extraction method for spatio-temporal data. GraFT uses a diffusion map to learn a graph over spatial pixels that enables for stochastic filtering of learned sparse representations over each pixel’s time-trace. The sparse representations are modeled as in a hierarchical dictionary learning framework with correlated decompositions over the graph. GitHub repo NWB tutorials Cite

GitHub Repo stars for GraFT