On secant defectiveness and identifiability of Segre-Veronese varieties (Q2104844)

The paper studies the non-defectivity of the higher secant varieties of the Segre-Varieties \(X\subset \mathbb{P}^N\) and the so-called uniqueness (or generic uniqueness): the uniqueness of a representation a general element of the \(h\)-secant variety of \(X\) as in a linear span of a unique subset of \(X\) with cardinality \(h\). Here \(X\cong \mathbb {P}^{n_1}\times \cdots \mathbb {P}^{n_r}\) embedded by the line bundle of multidegree \((d_1,\dots ,d_r)\). Generic uniqueness is stronger than non-defectivity, but, roughly speaking, non-defectivity for higher order secant varieties implies generic uniqueness. The authors give a positive for both problems (for a fixed \(n_1,\dots ,n_r,d_1,\dots,d_r\) any easy function \(a(n_1,\dots ,d_r)\) (resp. \(b(n_1,\dots ,d_r)\) such that for all \(h\le a(n_1,\dots ,d_r)\) (resp. \(h\le a(n_1,\dots ,d_r)\)) the \(h\)-secant variety is non-defectivitive (resp. it has generic uniqueness). A table shows how big is their improvements with respect to previous work. Their proof use convex geometry. There are many different competing tools. Very recently (solving a conjecture of \textit{F. Galuppi} and \textit{A. Oneto} [Adv. Math. 409 B, Article ID 108657, 58 p. (2022; Zbl 1503.14049)], proved the case \(r=2\) with the full list of the exceptional cases.