W. Feng and Jane W. S. Liu. An extended imprecise computation model for timeconstrained speech processing and generation. In Proceedings of the IEEE Workshop on Real-Time Applications, pages 76--80, New York, NY, May 1993.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....However, their generality is restricted by the requirement that valid results be available not only on completion but at all times. Hence, anytime and any dimension algorithms are also subsets of the more general class of multi fidelity algorithms. The third extension, imprecise computations [29, 30, 48], supports graceful degradation of real time systems under overload conditions. Each computation is modeled as a mandatory part followed by an optional anytime part that improves the precision of the result. The real time scheduler ensures that the mandatory portion of every task meets its ....

W. Feng and Jane W. S. Liu. An extended imprecise computation model for timeconstrained speech processing and generation. In Proceedings of the IEEE Workshop on Real-Time Applications, pages 76--80, New York, NY, May 1993.

....to execute additional processing levels. This strategy leads to a substantial quality gain not only over the baseline approach but also over all similar scheduling approaches that use a single duration such as the average duration [3] the most likely duration [11] or the worst case duration [2]. Furthermore, the baseline approach is based on a pessimistic strategy that discards all the levels that may violate the deadline while our approach takes risks when they are justified in terms of expected quality. This explains the substantially slower growth of the comprehensive quality of ....

W. Feng and J.W.S Liu, `An extended imprecise computation model for time-constrained speech processing and generation', in IEEE Workshop on Real-Time Applications, pp. 76--80, (1993).

....of a real time system introduces a problem of execution monitoring: determining how long to run. This kind of scheduling problems are common in real time AI applications, such as medical monitoring [Hayes Roth, 1990] real time data transmission [MillanLopez et al. 1994] speech processing [Feng and Liu, 1993], mobile robot navigation [Zilberstein, 1996] and flexible manufacturing. The real time data transmission application that we address in this paper is another example. These application require a meta level control mechanism that can deal with (1) dynamic environment, 2) limited amount of time to ....

W. Feng and J.W.S Liu. An extended imprecise computation model for timeconstrained speech processing and generation. In IEEE Workshop on Real-Time Applications, pages 76-80, 1993.

....Quality of Service improves. For example, allocating a tracking application more processor time to complete a better but more complex tracking algorithm leads to a lower probability 7 of producing false tracks. Research on Imprecise Computation in areas as diverse as digital control systems [5] and artificial intelligence [12] have shown it to be a feasible and effective approach. Under the model of Imprecise Computation, we logically divide the requirement of each flexible application into two parts: a mandatory part and an optional part. this division is done in such a way that the ....

....composed of a series of flexible tasks, which carry different parts of the responsibilities of the application, and may have their own specific Quality of Service demands. In order to model the QoS demands of these flexible applications or tasks, we apply the Imprecise Computation Model [4] 12] [5] [13] which originated from the research of hard real time scheduling for computational tasks. The essence of the Imprecise Computation Model is to make the scheduling of real time systems easier by dividing the processing of each task and its result in the system into two parts: a mandatory ....

W. Feng and J. W.S. Liu. An Extended Imprecise Computation Model for Time-Constrained Speech Processing and Generation. Proceedings of the IEEE Workshop on Real-Time Applications, May 1993.

....project (Contract DABT63 96 C 0044) out of n consecutive instances of a task. The techniques mentioned above tacitly assume that a task s output is of no value if it is not executed completely. However, in many application areas such as multimedia applications [17] image and speech processing [4, 6, 19], timedependent planning [3] robot control navigation systems [21] medical decision making [9] information gathering [7] realtime heuristic search [12] and database query processing [20] a partial or approximate but timely result is usually acceptable. The imprecise computation [5, 15] and ....

....executions. The first derivative of a nondecreasing concave function is nonincreasing. Linear and general concave functions are considered as the most realistic and typical in the literature since they adequately capture the behavior of many application areas like image and speech processing [4, 6, 19], time dependent planning [3] robot control navigation systems [21] real time heuristic search [12] information gathering [7] multimedia applications [17] and database query processing [20] In this paper, we show that the case of convex reward functions is an NP Hard problem and thus focus on ....

[Article contains additional citation context not shown here]

W. Feng and J. W.-S. Liu. An extended imprecise computation model for time-constrained speech processing and generation. In Proceedings of the IEEE Workshop on Real-Time Applications, May 1993.

....provide hard real time guarantees to k out of n consecutive instances of a task. The techniques mentioned above tacitly assume that a task s output is of no value if it is not executed completely. However, in many application areas such as multimedia applications [26] image and speech processing [5, 6, 9, 28], time dependent planning [4] robot control navigation systems [12, 30] medical decision making [13] information gathering [10] real time heuristic search [17] and database query processing [29] a partial or approximate but timely result is usually acceptable. The imprecise computation [7, ....

....in the output quality is obtained during the first portions of optional executions. Linear and general concave functions are considered as the most realistic and typical in the literature since they adequately capture the behavior of many application areas like image and speech processing [5, 6, 9, 28], multimedia applications [26] time dependent planning [4] robot control navigation systems [30] real time heuristic search [17] information gathering [10] and database query processing [29] In this paper, we show that the case of convex reward functions is an NP Hard problem and thus focus ....

[Article contains additional citation context not shown here]

W. Feng and J. W.-S. Liu. An extended imprecise computation model for time-constrained speech processing and generation. In Proceedings of the IEEE Workshop on Real-Time Applications, May 1993.

....Quality of Service improves. For example, allocating a tracking application more processor time to complete a better but more complex tracking algorithm leads to a lower probability 7 of producing false tracks. Research on Imprecise Computation in areas as diverse as digital control systems [5] and artificial intelligence [12] have shown it to be a feasible and effective approach. Under the model of Imprecise Computation, we logically divide the requirement of each flexible application into two parts: a mandatory part and an optional part. this division is done in such a way that the ....

....composed of a series of flexible tasks, which carry different parts of the responsibilities of the application, and may have their own specific Quality of Service demands. In order to model the QoS demands of these flexible applications or tasks, we apply the Imprecise Computation Model [4] 12] [5] [13] which originated from the research of hard real time scheduling for computational tasks. The essence of the Imprecise Computation Model is to make the scheduling of real time systems easier by dividing the processing of each task and its result in the system into two parts: a mandatory ....

W. Feng and J. W.S. Liu. An Extended Imprecise Computation Model for Time-Constrained Speech Processing and Generation. Proceedings of the IEEE Workshop on Real-Time Applications, May 1993.

....provide hard real time guarantees to k out of n consecutive instances of a task. The techniques mentioned above tacitly assume that a task s output is of no value if it is not executed completely. However, in many application areas such as multimedia applications [25] image and speech processing [5, 6, 9, 27], time dependent planning [4] robot control navigation systems [12, 29] medical decision making [13] information gathering [10] real time heuristic search [17] and database query processing [28] a partial or approximate but timely result is usually acceptable. The imprecise computation [7, ....

....in the output quality) is obtained during the first portions of optional executions. Linear and general concave functions are considered as the most realistic and typical in the literature since they adequately capture the behavior of many application areas like image and speech processing [5, 6, 9, 27], multimedia applications [25] time dependent planning [4] robot control navigation systems [29] real time heuristic search [17] information gathering [10] and database query processing [28] For completeness, in this paper, we show that the case of convex reward functions is an NP Hard ....

[Article contains additional citation context not shown here]

W. Feng and J. W.-S. Liu. An extended imprecise computation model for time-constrained speech processing and generation. In Proceedings of the IEEE Workshop on Real-Time Applications, pp. 76-80, May 1993.

....provide hard real time guarantees to k out of n consecutive instances of a task. The techniques mentioned above tacitly assume that a task s output is of no value if it is not executed completely. However, in many application areas such as multimedia applications [17] image and speech processing [4, 6, 19], timedependent planning [3] robot control navigation systems [21] medical decision making [9] information gathering [7] realtime heuristic search [12] and database query processing [20] a partial or approximate but timely result is usually acceptable. The Imprecise Computation [5, 15] and ....

....of optional executions. The first derivative of a nondecreasing concave function is nonincreasing. Linear and general concave functions are considered the most realistic and typical in the literature since they adequately capture the behavior of many application areas such as those mentioned above [4, 6, 19, 3, 21, 12, 7, 17, 20]. In this paper, we show that the case of convex reward functions is an NP Hard problem and thus focus on linear and concave reward functions. Reward functions with 0 1 constraints, where no reward is accrued unless the entire optional part is executed, or step reward functions have also received ....

[Article contains additional citation context not shown here]

W. Feng and J. W.-S. Liu. An extended imprecise computation model for time-constrained speech processing and generation. In Proceedings of the IEEE Workshop on Real-Time Applications, May 1993.

....Rather than have explicit per application brokers, I propose to have the system learn to be a broker, by monitoring the application. The Rialto operating system [15] has, in its design, the notion of using user preferences to guide resource allocations between applications. Imprecise computations [6, 7, 14], support graceful degradation of real time systems under overload conditions. Each computation is modeled as a mandatory part followed by an optional anytime part that improves the precision of the result. The amount of time spent in the optional part can be viewed as a measure of fidelity; the ....

W. Feng and Jane W. S. Liu. An Extended Imprecise Computation Model for Time-constrained Speech Processing and Generation. In Proceedings of the IEEE Workshop on Real-Time Applications, pages 76--80, New York, New York, May 1993.

....The consequences of missing a deadline in a hard real time environment may be serious, even catastrophic. On the aother hand, an approximate but timely result may be acceptable in many application areas. Examples of such applications are multimedia applications [15] image and speech processing [6, 8, 18], time dependent planning [5] robot control navigation systems [20] medical decision making [10] information gathering [9] real time heuristic search [11] and database query processing [19] The imprecise computation approach is a technique of improving the responsiveness and resource ....

....with mandatory parts and dynamic arrivals are described in [7] The notion of reward in the IRIS model is analogous to that of precision error in the Imprecise Computation model. Linear and general concave functions represent most of the real world applications such as those mentioned above [6, 8, 18, 15, 5, 20, 11, 9, 19]. Note that the first derivative of a nondecreasing concave function is nonincreasing. In this paper, we focus on linear and concave reward functions. Maximization of the total reward in a system of tasks with 0 1 constraints, where no reward is accrued for a partial execution was shown to be ....

W. Feng and J. W.-S. Liu. An extended imprecise computation model for time-constrained speech processing and generation. In Proceedings of the IEEE Workshop on Real-Time Applications, May 1993.

....However, the requirement that execution be interruptible at any point in time is an added constraint that restricts their generality. Hence, anytime and any dimension algorithms are also subsets of the more general class of multi fidelity algorithms. The third extension, imprecise computations [4, 5, 11], support graceful degradation of real time systems under overload conditions. Each computation is modeled as a mandatory part followed by an optional anytime part that improves the precision of the result. The real time scheduler ensures that the mandatory portion of every task meets its ....

Feng, W. and Liu, J. W. S. An Extended Imprecise Computation Model for Time-constrained Speech Processing and Generation. In Proceedings of the IEEE Workshop on RealTime Applications, pages 76--80, New York, New York, May 1993.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC