Search

Speech translation research has made significant progress over the years with many high-visibility efforts showing that translation of spontaneously spoken speech from and to diverse languages is possible and applicable in a variety of domains. As language and domains continue to expand, practical concerns such as portability and reconfigurability of speech come into play: system maintenance becomes a key issue and data is never sufficient to cover the changing domains over varying languages. In this paper, we discuss strategies to overcome the limits of today's speech translation systems. In the first part, we describe our layered system architecture that allows for easy component integration, resource sharing across components, comparison of alternative approaches, and the migration toward hybrid desktop/PDA or stand-alone PDA systems. In the second part, we show how flexibility and reconfigurability is implemented by more radically relying on learning approaches and use our English–Thai two-way speech translation system as a concrete example.

The increasing interest in multilingual applications like speech-to-speech translation systems is accompanied by the need for speech recognition front-ends in many languages that can also handle multiple input languages at the same time. In this paper we describe a universal speech recognition system that fulfills such needs. It is trained by sharing speech and text data across languages and thus reduces the number of parameters and overhead significantly at the cost of only slight accuracy loss. The final recognizer eases the burden of maintaining several monolingual engines, makes dedicated language identification obsolete and allows for code-switching within an utterance. To achieve these goals we developed new methods forconstructing multilingual acoustic models and multilingual n-gram language models.

This paper presents a context-dependent hybrid connectionist speech recognition system that uses a set of generalized hierarchical mixtures of experts (HME) to estimate context-dependent posterior acoustic class probabilities. The connectionist part of the system is organized in a modular fashion, allowing the distributed training of such a system on regular workstations. Context classes are based on polyphonic contexts, clustered using decision trees which we adopt from our continuous density HMM recognizer JANUS (Waibel et al., 1996). The system is evaluated on ESST, an English speaker-independent spontaneous speech database. Context dependent modeling is shown to yield significant improvements over simple context-independent modeling, requiring only small additional overhead in terms of training and decoding time.

This paper deals with buffer management for tree-based access-path structures in real-time database system kernels. A predictable and efficient buffer replacement strategy for B- and B/sup +/-trees is introduced. The strategy called LC (Level Control) uses the structural information known about the trees and the tree search to determine a node to be replaced. The basic idea is to use the level of a node as replacement criteria. This allows a certain prediction of buffer hits and misses for tree search operations, even if the item's value searched for is unknown. A small prediction interval can be calculated for search operations and a larger interval for update operations. Simulations show the correctness of these calculations and the efficiency of LC for B- and B/sup +/-trees compared to common other replacement strategies.

This paper presents fill algorithms for boundary-defined regions in raster graphics. The algorithms require only a constant-size working memory. The methods presented are based on the so-called “seed fill” algorithms that use the internal connectivity of the region with a given inner point. Basic methods, as well as additional hcuristics for speeding up the algorithm, are described and verified. Empirical results are used to compare the time complexities of the algorithms for different classes of regions.