Sunday, May 29, 2016



Soy un nuevo usuario

Olvidé mi contraseña

Entrada usuarios

Lógica Matemáticas Astronomía y Astrofísica Física Química Ciencias de la Vida
Ciencias de la Tierra y Espacio Ciencias Agrarias Ciencias Médicas Ciencias Tecnológicas Antropología Demografía
Ciencias Económicas Geografía Historia Ciencias Jurídicas y Derecho Lingüística Pedagogía
Ciencia Política Psicología Artes y Letras Sociología Ética Filosofía

rss_1.0 Recursos de colección

HKUST Institutional Repository (12,074 recursos)
Repository of Hong Kong University of Science and Technology. Managed by the HKUST Library.

ELEC Preprints

Mostrando recursos 1 - 6 de 6

1. Efficient motion estimation algorithms for resolution down-conversion of compressed video - Wong, Wai Chuen; Au, Oscar C.
In this paper, we propose several fast motion estimation algorithms for the compression of video down converted from some original compressed video. Firstly, we propose the Predictive Motion Estimation (PME) and its variants which achieve significantly higher PSNR than existing algorithms by perfoming a simple low-complexity search using candidate motion vectors from the original compressed video. Secondly, we propose the modified PME (MPME) with increased efficiency for cases with high conversion factor. Thirdly, we propose the ME-SVF which can achieve significantly higher PSNR than existing algorithtms by applying spatial varying filtering on the motion field without performing any search.

2. Sub-phonetic polynomial segment model for large vocabulary continuous speech recognition - Au Yeung, Siu-Kei; Li, Chak Fai; Siu, Man Hung
Polynomial Segment Model (PSM) has opened up an alternative research direction for acoustic modeling. In our previous papers~\cite{ref:Fai1,ref:Fai2}, we proposed efficient incremental likelihood evaluation and EM training algorithms for PSM, making it possible to train and recognize using PSM alone. In this paper, we shift our focus to make it feasible to use PSM on large vocabulary recognition. First, we used sub-phonetic PSM that represents a phoneme as multiple independent segmental units. Second, we derived and compared different top-down mixture growing approaches that are orders of magnitude more efficient than previously proposed agglomerative clustering techniques. Experimental results show that the...

3. Optimal antenna selection based on capacity maximization for MIMO systems in correlated channels - Dai, Lin; Sfar, Sana; Letaief, K. B.
Recent work has shown that multiple-input multiple-output (MIMO) systems with multiple antennas at both the transmitter and receiver are able to achieve great capacity improvement. In such systems, it is desirable to select a subset of the available antennas so as to reduce the number of RF chains. This paper addresses the problem of optimal antenna selection in correlated channels. We consider a narrow-band communication system with M transmit and N receive antennas. We present the criterion for selecting the optimal Lt out of M transmit and Lr out of N receive antennas in terms of capacity maximization, assuming that...

4. A quasi-orthogonal group space-time architecture to achieve a better diversity-multiplexing tradeoff - Dai, Lin; Sfar, Sana; Letaief, K. B.
Most existing MIMO (Multiput-Input Multiput-Output) schemes optimize only either the diversity gain or the multiplexing gain. To obtain a good tradeoff between these two, the Quasi-Orthogonal Group Space-Time (QoGST) architecture is proposed, wherein the transmit stream is subgrouped but encoded via an inter-group space-time block encoder, with group interference suppression at the receiver. This paper also considers another combined space-time coding and layered space-time architecture, which we refer to as Group Layered Space-Time (GLST), where space-time block coding is employed within each group. Under the assumption of Rayleigh fading and a prior perfect channel state information at the receiver, a...

5. An efficient detector for combined space-time coding and layered processing - Dai, Lin; Sfar, Sana; Letaief, K. B.
Group Layered Space-time architecture (GLST) combines space-time block coding and layered space-time processing, where the transmit stream is partitioned into different groups and in each group space-time block coding is applied. In the traditional receiver of GLST, group detection is applied first so as to suppress the interference from other groups and then decoding is performed for the desired group. In this paper, a novel detector is proposed, in which the entire groups are decoded first and then group detection is performed next. Theoretical analysis will demonstrate that the new detector can achieve a significant capacity gain compared to the...

6. CoSAM2 : a unified control system architecture for multifingered manipulation - Li, Zexiang; Qin, Z.; Jiang, S.
In this paper, we describe a unified Control System Architecture for Multifingered Manipulation (CoSAM2). CoSAM2 achieves simultaneously three objectives of multifingered manipulation: (a) Motion trajectory (velocity/force) tracking of a grasped object; (b) Improving the grasp configuration in the course of object manipulation; and (c) Optimizing grasping forces to enforce contact constraint and compensate for external object wrenches. CoSAM2 is organized in a modular and hierarchic structure so that each module implements a specified function using inputs from its predecessors and a minimum number of sensory data signals. CoSAM2 is also flexible in accommodating addition of new modules such as the...


Busque un recurso