FFTKF applies an FFT low-pass mask and a Kalman filter to privatized gradients, reporting modest accuracy gains over DP-SGD and DiSK on several image benchmarks, but with incomplete privacy accounting.
Brendan McMahan, and Swaroop Ramaswamy
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Fast Fourier Transform-Based Spectral and Temporal Gradient Filtering for Differential Privacy
FFTKF applies an FFT low-pass mask and a Kalman filter to privatized gradients, reporting modest accuracy gains over DP-SGD and DiSK on several image benchmarks, but with incomplete privacy accounting.