Spatial Neighboring Scattering Transform: A Cross-Channel Amplitude Coupling Measure for EEG Connectivity
arXiv:2607.08855
2026
Architecture
1 ideas extracted · analyzed Aug 30, 2026
What the math gives to ML
The paper introduces a concrete multichannel generalization of wavelet scattering: instead of applying a modulus independently to each channel, it forms a complex cross-channel product and then takes its magnitude. This produces a phase-robust amplitude-envelope interaction feature, retaining band-limited temporal structure while being less sensitive to time shifts than raw cross-channel products. The most promising neural-network transfer is a differentiable multichannel scattering front end or auxiliary representation loss that exposes local cross-channel amplitude coupling to EEG, audio, or sensor-fusion models without requiring the network to discover this interaction from limited data.
Ideas from this paper
Unverified
2026
Add a differentiable SNST layer before an EEG classifier or sequence model. For every local channel pair and wavelet band, compute the magnitude of the complex cross-channel analytic response, then average it over a controllable temporal window and concatenate it with ordinary channelwise features. This gives the model an explicit, phase-robust amplitude-coupling representation that is especially useful when labeled training data are scarce.
Useful6/10
Difficulty4/10
Novelty7/10