Existing zero-shot learning (ZSL) methods usually learn a projection function between a feature space and a semantic embedding space(text or attribute space) in the training seen classes or testing unseen classes. However, the projection function cannot be used between the feature space and multi-semantic embedding spaces, which have the diversity characteristic for describing the different sem...

Introducing suitable features in the scribble-based foreground-background (Fg/Bg) segmentation problem is crucial. In many cases, the object of interest has different regions with different color modalities. The same applies to a non-uniform background. Fg/Bg color modalities can even overlap when the appearance is solely modeled using color spaces like RGB or Lab. In this paper, we purposefull...

Recently,manyworks have focused on the characterization of nonlinear dimensionality reductionmethods obtained by quantizing linear embeddings, e.g. to reach fast processing time, efficient data compression procedures, novel geometry-preserving embeddings or to estimate the information/bits stored in this reduced data representation. In this work, we prove that many linear maps known to respect ...

We consider the problem of embedding a graph on n vertices in Euclidean space R for k n Typically k would be or By posing the problem as minimising the squared norm of the appropriately weighted distance be tween adjacent points subject to natural normalising conditions we arrive at a formulation of the problem for which the optimal solution can be sim ply computed in terms of the eigenvectors ...

Predicting future behavior of chaotic time series system is a challenging area in the literature of nonlinear systems. The prediction's accuracy of chaotic time series is extremely dependent on the model and the learning algorithm. On the other hand the cyclic solar activity as one of the natural chaotic systems has significant effects on earth, climate, satellites and space missions. Several m...

We replace the usual Convex Integration formula by a Corrugation Process and introduce notion of Kuiper differential relations. This provides natural framework for construction solutions with self-similarity properties. consider case totally real relation, we prove that it is state isometric embedding theorem. then show embeddings obtained exhibit property. relations also enable uniform express...

The utilization of millimeter-wave (mmWave) bands in 5G networks poses new challenges to network planning. Vulnerability blockages at mmWave can cause coverage holes (CHs) the radio environment, leading link failure when a user enters these CHs. Detection CHs carries critical importance so that necessary remedies be introduced improve coverage. In this letter, we propose novel approach identify...

We present a new class of explicit marching schemes for the wave equation in complex geometry. They rely on a simple embedding of the domain in a uniformCartesian grid, which allows for efficient and automatic implementation but creates irregular cells near the boundary. While existing explicit finite difference schemes are generally restricted in the size of the time step that can be taken by ...

Subspace clustering groups data into several lowrank subspaces. In this paper, we propose a theoretical framework to analyze a popular optimization-based algorithm, Sparse Subspace Clustering (SSC), when the data dimension is compressed via some random projection algorithms. We show SSC provably succeeds if the random projection is a subspace embedding, which includes random Gaussian projection...

We address the problem of giving a clean and uniform mathematical model for handling user deened data types in imperative languages, contrary to the ad-hoc treatment usual in classical denotational semantics. The problem is solved by deen-ing the store as a homomorphic mapping of an algebraic structure of left values modelling containers into another one of right values modelling contents. Cons...