🤖 AI Summary
Unsupervised separation of multiscale seismic sources—such as second-scale “glitches” and tens-of-minutes atmospheric noise—in NASA’s InSight Mars seismic data remains challenging under strictly unsupervised conditions: no prior source knowledge and only mixed observations available.
Method: We propose a wavelet scattering spectrum–driven Nested Factor-VAE framework, introducing the first multiscale nested unsupervised separation paradigm that overcomes fixed-window limitations, enabling cross-scale source identification and probabilistic clustering. The wavelet scattering spectrum constructs an interpretable, low-dimensional, non-Gaussian stochastic process representation space.
Results: Evaluated on full-task InSight data, our method successfully disentangles heterogeneous sources—including glitches and atmospheric noise—with significantly higher separation accuracy than single-scale baselines. This enables independent characterization of Martian interior structure and surface environmental activity, advancing planetary seismology through robust, physics-informed unsupervised learning.
📝 Abstract
Unsupervised source separation involves unraveling an unknown set of source signals recorded through a mixing operator, with limited prior knowledge about the sources, and only access to a dataset of signal mixtures. This problem is inherently ill-posed and is further challenged by the variety of timescales exhibited by sources in time series data from planetary space missions. As such, a systematic multi-scale unsupervised approach is needed to identify and separate sources at different timescales. Existing methods typically rely on a preselected window size that determines their operating timescale, limiting their capacity to handle multi-scale sources. To address this issue, we propose an unsupervised multi-scale clustering and source separation framework by leveraging wavelet scattering spectra that provide a low-dimensional representation of stochastic processes, capable of distinguishing between different non-Gaussian stochastic processes. Nested within this representation space, we develop a factorial variational autoencoder that is trained to probabilistically cluster sources at different timescales. To perform source separation, we use samples from clusters at multiple timescales obtained via the factorial variational autoencoder as prior information and formulate an optimization problem in the wavelet scattering spectra representation space. When applied to the entire seismic dataset recorded during the NASA InSight mission on Mars, containing sources varying greatly in timescale, our approach disentangles such different sources, e.g., minute-long transient one-sided pulses (known as"glitches") and structured ambient noises resulting from atmospheric activities that typically last for tens of minutes, and provides an opportunity to conduct further investigations into the isolated sources.
