Augmented listening devices such as hearing aids often perform poorly in noisy and reverberant environments with many competing sound sources. Large distributed microphone arrays can improve performance, but data from...
详细信息
ISBN:
(数字)9781728155494
ISBN:
(纸本)9781728155500
Augmented listening devices such as hearing aids often perform poorly in noisy and reverberant environments with many competing sound sources. Large distributed microphone arrays can improve performance, but data from remote microphones often cannot be used for delay-constrained real-time processing. We present a cooperative audio source separation and enhancement system that leverages wearable listening devices and other microphone arrays spread around a room. The full distributed array is used to separate sound sources and estimate their statistics. Each listening device uses these statistics to design real-time binaural audio enhancement filters using its own local microphones. The system is demonstrated experimentally using 10 speech sources and 160 microphones in a large, reverberant room.
Clustering with accurate results have become a topic of high interest. Dirichlet Process Mixture (DPM) is a model used for clustering with the advantage of discovering the number of clusters automatically. It is highl...
详细信息
Malleability is the property of an application to be dynamically rescaled at run time. It requires the possibility to dynamically add or remove resources to the infrastructure without interruption. Yet, many Big Data ...
详细信息
ISBN:
(纸本)9781728101927
Malleability is the property of an application to be dynamically rescaled at run time. It requires the possibility to dynamically add or remove resources to the infrastructure without interruption. Yet, many Big Data applications cannot benefit from their inherent malleability, since their colocated distributed storage system is not malleable in practice. Commissioning or decommissioning storage nodes is generally assumed to be slow, as such operations have typically been designed for maintenance only. New technologies, however, enable faster data transfers. Still, evaluating the performance of rescaling operations on a given platform is a challenge in itself: no tool currently exists for this purpose. We introduce Puffer bench, a benchmark for evaluating how fast one can scale up and down a distributed storage system on a given infrastructure and, thereby, how viably can one implement storage malleability on it. Besides, it can serve to quickly prototype and evaluate mechanisms for malleability in existing distributed storage systems. We validate Pufferbench against theoretical lower bounds for commission and decommission: it can achieve performance within 16% of them. We use Puffer bench to evaluate in practice these operations in HDFS: commission in HDFS could be accelerated by as much as 14 times! Our results show that: (1) the lower bounds for commission and decommission times we previously established are sound and can be approached in practice;(2) HDFS could handle these operations much more efficiently;most importantly, (3) malleability in distributed storage systems is viable and should be further leveraged for Big Data applications.
Complex event processing (CEP) systems continuously process input event streams to detect patterns. Over time, the input event rate might fluctuate and overshoot the system's capabilities. One way to reduce the ov...
详细信息
ISBN:
(纸本)9781728108582
Complex event processing (CEP) systems continuously process input event streams to detect patterns. Over time, the input event rate might fluctuate and overshoot the system's capabilities. One way to reduce the overload on the system is to use load shedding. In this paper, we propose a load shedding strategy for CEP systems which drops a portion of the CEP operator's internal state (a.k.a. partial matches) to maintain a given latency bound. The crucial question here is how many and which partial matches to drop so that a given latency bound is maintained while minimizing the degradation in the quality of results. In the stream processing domain, different load shedding strategies have been proposed that mainly depend on the importance of individual tuples. However, as CEP systems perform pattern detection, the importance of events is also influenced by other events in the stream. Our load shedding strategy uses Markov chain and Markov reward process to predict the utility/importance of partial matches to determine the ones to be dropped. In addition, we represent the utility in a way that minimizes the overhead of load shedding. Furthermore, we provide algorithms to decide when to start dropping partial matches and how many partial matches to drop. By extensively evaluating our approach on three real-world datasets and several representative queries, we show that the adverse impact of our load shedding strategy on the quality of results is considerably less than the impact of state-of-the-art load shedding strategies.
Downtime is a key performance index for industrial automation systems. An industrial automation system achieves maximum productivity when its downtime is reduced to the minimum. One approach to minimize downtime is to...
详细信息
Downtime is a key performance index for industrial automation systems. An industrial automation system achieves maximum productivity when its downtime is reduced to the minimum. One approach to minimize downtime is to predict system faults and recover from them automatically. A cloud-based decision support system is proposed for rapid problem identifications and to assist the self-management processes. By running multiple parallel simulations of control software with real-time inputs ahead of system time, faults could be detected and corrected automatically using autonomous industrial software agents. Fault trees, as well as control algorithms, are modeled using IEC 61499 function blocks that can be directly executed on both physical controllers and cloud services. A case study of water heating process is used to demonstrate the self-healing process supported by the cloud-based decision support system.
On battery-operated devices, energy and power consumption are main concerns. With the recent advancement of technology, mobile devices can be integrated with traditional systems for running complex computations. In fa...
详细信息
ISBN:
(纸本)9781538650486
On battery-operated devices, energy and power consumption are main concerns. With the recent advancement of technology, mobile devices can be integrated with traditional systems for running complex computations. In fact, mobile devices can easily become part of computational networks and share their computational and memory resources. Despite this, traditional simulation frameworks are not designed to perform well on heterogeneous networks. This is mainly due to the limited computational resources that are available on mobile devices. In this paper, we propose SEECSSim (SEECSSim is derived from School of Electrical Engineering and Computer Science (SEECS)) that is a simulation framework specifically designed for mobile devices. SEECSSim includes state-of-theart distributed synchronization algorithms that are implemented to run on mobile or embedded devices. To benchmark the proposed framework, the well-known PHOLD model is used and performance results are reported in terms of execution time, CPU usage, memory and energy consumption.
Volunteer computing (VC) or distributed computing projects are common in the citizen cyberscience (CCS) community and present extensive opportunities for scientists to make use of computing power donated by volunteers...
详细信息
Volunteer computing (VC) or distributed computing projects are common in the citizen cyberscience (CCS) community and present extensive opportunities for scientists to make use of computing power donated by volunteers to undertake large-scale scientific computing tasks. VC is generally a noninteractive process for those contributing computing resources to a project, whereas volunteer thinking (VT) or distributed thinking allows volunteers to participate interactively in CCS projects to solve human computation tasks. In this paper, we describe the integration of three tools, the Virtual Atom Smasher (VAS) game developed by CERN, LiveQ, a job distribution middleware, and CitizenGrid, an online platform for hosting and providing computation to CCS projects. This integration demonstrates the combining of VC and VT to help address the scientific and educational goals of games like VAS. This paper introduces the three tools and provides details of the integration process along with further potential usage scenarios for the resulting platform.
This work presents an all-hardware realtime implementation of the SIFT algorithm. The implementation exploits pipeline structures both in the keypoint extraction and in the descriptor generation stages to achieve rea...
详细信息
ISBN:
(纸本)9781450365116
This work presents an all-hardware realtime implementation of the SIFT algorithm. The implementation exploits pipeline structures both in the keypoint extraction and in the descriptor generation stages to achieve realtime requirements. To allow a feasible hardware implementations, some simplifications to the original algorithm have been required. The architecture has been synthesized on a Xilinx FPGA. It generates 3072 descriptor vectors for VGA images at 99 frames per second.
暂无评论