Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper analyzes the non-convex optimization landscape of low-tubal-rank tensor sensing under a tubal restricted isometry property using balanced factorizations. The authors prove a benign strict-saddle landscape with zero spurious local minima for arbitrary Fourier multi-rank profiles, while showing that the actual local geometry is dictated by slice-wise Fourier ranks rather than nominal tubal rank. Crucially, nonuniform Fourier profiles induce quartically flat directions via hidden frequency-wise overparameterization, even when factor widths are tuned exactly to the ground-truth tubal rank.
Perfectly rank-matched tensor factorizations still suffer from quartically flat optimization valleys whenever the underlying Fourier slice ranks are nonuniform.
We study the optimization landscape of low-tubal-rank tensor sensing through a balanced factorization. Under a tubal restricted isometry condition, we establish a quantitative strict-saddle landscape with no spurious local minima for arbitrary Fourier multi-rank profiles. We further show that the local geometry depends on the Fourier-slice ranks rather than the tubal rank alone. Uniform ranks yield quadratic growth transverse to the solution orbit, whereas nonuniform ranks produce quartically flat directions through hidden frequency-wise overparameterization, even when the factor width equals the exact tubal rank. Numerical experiments illustrate the global optimization behavior and the contrasting local geometries.