We study a framework of multi-target state initialization employed over a tree-like sensor network. Such network provides graceful aggregation of target state statistics by exploiting limited bandwidth, and makes it possible to approximate the target states based on particle filtering defined in the context of Random-FiniteSet (RFS) theory. This contrasts with conventional multi-target filterin...

We show that for all given n, t, w ∈ {1, 2, . . .} with n < 2, an array of n entries of w bits each can be represented on a word RAM with a word length of w bits in at most nw + dn(t/(2w))e bits of uninitialized memory to support constant-time initialization of the whole array and O(t)-time reading and writing of individual array entries. At one end of this tradeoff, we achieve initialization a...

Object-oriented class-based languages provide mechanisms for the initialization of newly created objects. These mechanisms specify how an object is initialized and what information is needed to do so. The initialization protocol is usually implemented as a list of constructors. It is often the case that the initialization protocol concerns some orthogonal properties of objects. Unfortunately, i...

Word embeddings are the interface between the world of discrete units of text processing and the continuous, differentiable world of neural networks. In this work, we examine various random and pretrained initialization methods for embeddings used in deep networks and their effect on the performance on four NLP tasks with both recurrent and convolutional architectures. We confirm that pretraine...

Saarinen recently proposed a chosen IV statistical attack, called the d-monomial test, and used it to find weaknesses in several proposed stream ciphers. In this paper we generalize this idea and propose a framework for chosen IV statistical attacks using a polynomial description. We propose a few new statistical attacks, apply them on some existing stream cipher proposals, and give some conclu...

This work presents a robust approach for fast initialization of an Active Appearance Model for subsequent segmentation of cardiac MRI data. The method automatically determines AAM initialization parameters: position, orientation, and scaling of the model. Four steps are carried out: (1) variance images over time are calculated to find a bounding box that roughly defines the heart region; (2) ci...

To analyze the applicability of direct insertion of total suspended matter (TSM) concentration field based on turbidity derived from satellite data to numerical simulation, dispersion studies of suspended matter in Lake Säkylän Pyhäjärvi (lake area 154 km²; mean depth 5.4 m) were conducted using the 3D COHERENS simulation model. To evaluate the practicality of direct insertion, five cases with ...