BZPEERPapers, people, evidence and research pathsDiscover
Research tools
Network
Opportunities
Account
Source author record
Shrawan Kumar appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.
Catalog footprint
Research graph
Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.
BZPEER is loading the nearby papers, people, topics and institutions for this page.
Published work
Lusztig defined an abelian category ${\mathscr{C}}_{k}$ of a class of representations of a multi-loop algebra and asked various questions connecting it to the modular representation theory of simple algebraic groups in char. p>0. The aim of this paper is to show that some of these questions have negative answer.
Following some work of Aluffi-Mihalcea-Schürmann-Su for the CSM classes of Schubert cells and some elaborate computer calculations by R. Rimanyi and L. Mihalcea, I conjecture that the CSM classes of the Richardson cells expressed in the Schubert basis have nonnegative coefficients. This conjecture was principally motivated by a new product $\square$ coming from the Segre classes in the cohomology of flag varieties (such that the associated Gr of this product is the standard cup product) and the conjecture that the structure constants of this new product $\square$ in the standard Schubert basis have alternating sign behavior. I prove that this conjecture on the sign of the structure constants of $\square$ would follow from my above positivity conjecture about the CSM classes of Richardson cells.
The purpose of this work is, to provide a YOLOv5 deep learning-based social distance monitoring framework using an overhead view perspective. In addition, we have developed a custom defined model YOLOv5 modified CSP (Cross Stage Partial Network) and assessed the performance on COCO and Visdrone dataset with and without transfer learning. Our findings show that the developed model successfully identifies the individual who violates the social distances. The accuracy of 81.7% for the modified bottleneck CSP without transfer learning is observed on COCO dataset after training the model for 300 epochs whereas for the same epochs, the default YOLOv5 model is attaining 80.1% accuracy with transfer learning. This shows an improvement in accuracy by our modified bottleneck CSP model. For the Visdrone dataset, we are able to achieve an accuracy of upto 56.5% for certain classes and especially an accuracy of 40% for people and pedestrians with transfer learning using the default YOLOv5s model for 30 epochs. While the modified bottleneck CSP is able to perform slightly better than the default model with an accuracy score of upto 58.1% for certain classes and an accuracy of ~40.4% for people and pedestrians.
For any reductive group $G$ and a parabolic subgroup $P$ with its Levi subgroup $L$, the first author [Ku2] introduced a ring homomorphism $ ξ^P_λ: Rep^\mathbb{C}_{λ-poly}(L) \to H^*(G/P, \mathbb{C})$, where $ Rep^\mathbb{C}_{λ-poly}(L)$ is a certain subring of the complexified representation ring of $L$ (depending upon the choice of an irreducible representation $V(λ)$ of $G$ with highest weight $λ$). In this paper we study this homomorphism for $G=SO(2n)$ and its maximal parabolic subgroups $P_{n-k}$ for any $2\leq k\leq n-1$ (with the choice of $V(λ) $ to be the defining representation $V(ω_1) $ in $\mathbb{C}^{2n}$). Thus, we obtain a $\mathbb{C}$-algebra homomorphism $ ξ_{n,k}^D: Rep^\mathbb{C}_{ω_1-poly}(SO(2k)) \to H^*(OG(n-k, 2n), \mathbb{C})$. We determine this homomorphism explicitly in the paper. We further analyze the behavior of $ ξ_{n,k}^D$ when $n$ tends to $\infty$ keeping $k$ fixed and show that $ ξ_{n,k}$ becomes injective in the limit. We also determine explicitly (via some computer calculation) the homomorphism $ ξ^P_λ$ for all the exceptional groups $G$ (with a specific `minimal' choice of $λ$) and all their maximal parabolic subgroups except $E_8$.
Let $\mathfrak{g}$ be a reductive Lie algebra and let $\vec{V}(\vecλ)$ be a tensor product of $k$ copies of finite dimensional irreducible $\mathfrak{g}$-modules. Choosing $k$ points in $\mathbb{C}$, $\vec{V}(\vecλ)$ acquires a natural structure of the current algebra $\mathfrak{g}\otimes \mathbb{C}[t]$-module. Following a work of Rao [R], we produce an explicit and complete set of $\mathfrak{g}$-module intertwiners of $\vec{V}(\vecλ)$ in terms of the action of the current algebra.
We prove a positivity result for the T-equivariant K-theory of flag varieties associated to any symmetrizable Kac-Moody group.
Kostant asked the following question: Let $\mathfrak{g}$ be a simple Lie algebra over the complex numbers. Let $λ$ be a dominant integral weight. Then, $V(λ)$ is a component of $ V(ρ)\otimes V(ρ)$ if and only if $λ\leq 2ρ$ under the usual Bruhat-Chevalley order on the set of weights. We give an affirmative answer to this question up to a saturation factor. In particular, the question is answered affirmatively for the special linear Lie algebras due to the Saturation Theorem of Knutson-Tao.
Although slicing removes code which has no bearing on property checking. However even after that, our study has found that there are predicates in program which have no bearing on property validation, although slicing could not eliminate them. We have cope up with a criteria to identify such predicates and then give a process to leverage them in scale up of property checking.
Recall the classical result that the cup product structure constants for the singular cohomology with integral coefficients of the Grassmannian of r-planes coincide with the Littlewood-Richardson tensor product structure constants for GL(r). Specifically, the result asserts that there is an explicit ring homomorphism ϕ: \Rep _{\poly}(GL(r)) to H^*(Gr(r, n)), where Gr(r, n) denotes the Grassmannian of r-planes in C^n and \Rep_{\poly} (GL(r)) denotes the polynomial representation ring of GL(r). This work seeks to achieve one possible generalization of this classical result for GL(r) and the Grassmannian Gr(r,n) to the Levi subgroups of any reductive group G and the corresponding flag varieties.
We show the Alon-Tarsi conjecture on Latin squares is equivalent to a very special case of a conjecture made independently by Hadamard and Howe, and to the non-vanishing of some interesting integrals over SU(n). Our investigations were motivated by geometric complexity theory.
Backward slicing has been used extensively in program understanding, debugging and scaling up of program analysis. For large programs, the size of the conventional backward slice is about 25% of the program size. This may be too large to be useful. Our investigations reveal that in general, the size of a slice is influenced more by computations governing the control flow reaching the slicing criterion than by the computations governing the values relevant to the slicing criterion. We distinguish between the two by defining data slices and control slices both of which are smaller than the conventional slices which can be obtained by combining the two. This is useful because for many applications, the individual data or control slices are sufficient. Our experiments show that for more than 50% of cases, the data slice is smaller than 10% of the program in size. Besides, the time to compute data or control slice is comparable to that for computing the conventional slice.
Let $\fg$ be a symmetrizable Kac-Moody Lie algebra with the standard Cartan subalgebra $\fh$ and the Weyl group $W$. Let $P_+$ be the set of dominant integral weights. For $λ\in P_+$, let $L(λ)$ be the irreducible, integrable, highest weight representation of $\fg$ with highest weight $λ$. For a positive integer $s$, define the {\em saturated tensor semigroup} as \begin{align*} Γ_s:= \{(λ_1, \dots, λ_s,μ)\in P_+^{s+1}: \exists\, N>1 \,\,\text{with}\,\, L(Nμ)\subset L(Nλ_1)\otimes \dots \otimes L(Nλ_s)\}. \end{align*} The aim of this paper is to begin a systematic study of $Γ_s$ in the infinite dimensional symmetrizable Kac-Moody case. In this paper, we produce a set of necessary inequalities satisfied by $Γ_s$. We further prove that any integer $d>0$ is a saturation factor for $A^{(1)}_1$ and 4 is a saturation factor for $A^{(2)}_2$.
The classical Hermitian eigenvalue problem addresses the following question: What are the possible eigenvalues of the sum A+B of two Hermitian matrices A and B, provided we fix the eigenvalues of A and B. A systematic study of this problem was initiated by H. Weyl (1912). By virtue of contributions from a long list of mathematicians, notably Weyl (1912), Horn (1962), Klyachko (1998) and Knutson-Tao (1999), the problem is finally settled. The solution asserts that the eigenvalues of A+B are given in terms of certain system of linear inequalities in the eigenvalues of A and B. These inequalities are given explicitly in terms of certain triples of Schubert classes in the singular cohomology of Grassmannians and the standard cup product. Belkale (2001) gave an optimal set of inequalities for the problem in this case. The Hermitian eigenvalue problem has been extended by Berenstein-Sjamaar (2000) and Kapovich-Leeb-Millson (2005) for any semisimple complex algebraic group G. Their solution is again in terms of a system of linear inequalities obtained from certain triples of Schubert classes in the singular cohomology of the partial flag varieties G/P (P being a maximal parabolic subgroup) and the standard cup product. However, their solution is far from being optimal. In a joint work with P. Belkale, we define a deformation of the cup product in the cohomology of G/P and use this new product to generate our system of inequalities which solves the problem for any G optimally (as shown by Ressayre). This article is a survey (with more or less complete proofs) of this additive eigenvalue problem.
Let G be a connected, adjoint, simple algebraic group over the complex numbers with a maximal torus T and a Borel subgroup B containing T. The study of zero weight spaces in irreducible representations of G has been a topic of considerable interest; there are many works which study the zero weight space as a representation space for the Weyl group. In this paper, we study the variation on the dimension of the zero weight space as the irreducible representation varies over the set of dominant integral weights for T which are lattice points in a certain polyhedral cone. The theorem proved here asserts that the zero weight spaces have dimensions which are piecewise polynomial functions on the polyhedral cone of dominant integral weights.
Let $\g$ be any simple Lie algebra over $\mathbb{C}$. Recall that there exists an embedding of $\mathfrak{sl}_2$ into $\g$, called a principal TDS, passing through a principal nilpotent element of $\g$ and uniquely determined up to conjugation. Moreover, $\wedge (\g^*)^\g$ is freely generated (in the super-graded sense) by primitive elements $ω_1, \dots, ω_\ell$, where $\ell$ is the rank of $\g$. N. Hitchin conjectured that for any primitive element $ω\in \wedge^d (\g^*)^\g$, there exists an irreducible $\mathfrak{sl}_2$-submodule $V_ω\subset \g$ of dimension $d$ such that $ω$ is non-zero on the line $\wedge^d (V_ω)$. We prove that the validity of this conjecture for simple simply-laced Lie algebras implies its validity for any simple Lie algebra. Let G be a connected, simply-connected, simple, simply-laced algebraic group and let $σ$ be a diagram automorphism of G with fixed subgroup K. Then, we show that the restriction map R(G) \to R(K) is surjective, where R denotes the representation ring over $\mathbb{Z}$. As a corollary, we show that the restriction map in the singular cohomology H^*(G)\to H^*(K) is surjective. Our proof of the reduction of Hitchin's conjecture to the simply-laced case relies on this cohomological surjectivity.
We construct deformations of the small quantum cohomology rings of homogeneous spaces G/P, and obtain an irredundant set of inequalities determining the multiplicative eigenvalue problem for the compact form K of G.
We look at the following question raised by Kollár and Peskine. (Actually, it is a slightly weaker version of their question.) Let $V_t$ be a family of rank two vector bundles on $\Bbb P^3$. Assume that the general member of the family is a trivial vector bundle. Then, is the special member $V_0$ also a trivial vector bundle? We show that this question is equivalent to the nonexistence of morphisms from $\Bbb P^3\to \mathcal{X}$, where $\mathcal{X}$ is the infinite Grassmannian associated to SL(2). We further reduce this question to the nonexistence of $\Bbb C^*$-equivariant morphisms from $\Bbb C^3\setminus \{0\} \to \mathcal{M}_d$ (for any $d>0$), where $\mathcal{M}_d$ is the Donaldson moduli space of isomorphism classes of rank two vector bundles $\mathcal{V}$ over $\Bbb P^2$ with trivial determinant and with second Chern class $d$ together with a trivialization of $\mathcal{V}_{|\Bbb P^1}$.
Let $X^v_w$ be a Richardson variety in the full flag variety $X$ associated to a symmetrizable Kac-Moody group $G$. Recall that $X^v_w$ is the intersection of the finite dimensional Schubert variety $X_w$ with the finite codimensional opposite Schubert variety $X^v$. We give an explicit $\bQ$-divisor $Δ$ on $X^v_w$ and prove that the pair $(X^v_w, Δ)$ has Kawamata log terminal singularities. In fact, $-K_{X^v_w} - Δ$ is ample, which additionally proves that $(X^v_w, Δ)$ is log Fano. We first give a proof of our result in the finite case (i.e., in the case when $G$ is a finite dimensional semisimple group) by a careful analysis of an explicit resolution of singularities of $X^v_w$ (similar to the BSDH resolution of the Schubert varieties). In the general Kac-Moody case, in the absence of an explicit resolution of $X^v_w$ as above, we give a proof that relies on the Frobenius splitting methods. In particular, we use Mathieu's result asserting that the Richardson varieties are Frobenius split, and combine it with a result of N. Hara and K.-I. Watanabe relating Frobenius splittings with log canonical singularities.
We give an explicit affine algebraic variety whose coordinate ring is isomorphic (as an algebra with the action of the Weyl group) with the equivariant cohomology of some Springer fibers.
We prove a generalization of Fulton's conjecture which relates intersection theory on an arbitrary flag variety to invariant theory.
We prove that the orbit closure of the determinant is not normal. A similar result is obtained for the orbit closure of the permanent multiplied by a power of a linear form.
Let G be a connected semisimple complex algebraic group and let P be a parabolic subgroup. In this paper we define a new (commutative and associative) product on the cohomology of the homogenous spaces G/P and use this to give a more efficient solution of the eigenvalue problem and also for the problem of determining the existence of G-invariants in the tensor product of irreducible representations of G. On the other hand, we show that this new product is intimately connected with the Lie algebra cohomology of the nil-radical of P via some works of Kostant and Kumar. We also initiate a uniform study of the geometric Horn problem for an arbitrary group $G$ by obtaining two (a priori) different sets of necessary recursive conditions to determine when a cohomology product of Schubert classes in G/P is non-zero. Hitherto, this was studied largely only for the group \SL(n).
We give a criterion for smoothness of a point in any Schubert variety in any G/B in terms of the nil Hecke ring.