This paper presents a graph-based approach for spoken term detection. Each first-pass retrieved utterance is a node on a graph and the edge between two ...

ABSTRACT. This paper presents a graph-based approach for spoken term detec- tion. Each first-pass retrieved utterance is a node on a graph and.

PDF | This paper presents a graph-based approach for spoken term detection. Each first-pass retrieved utterance is a node on a graph and the edge.

Experimental results show that this new approach offers significantly better performance than the previously proposed pseudo-relevance feedback approach, ...

ABSTRACT. This paper presents a graph-based approach for spoken term detec- tion. Each first-pass retrieved utterance is a node on a graph and.

➢Using graph-based re-ranking to improve spoken term detection with acoustic similarity. ➢With MLLR acoustic model, MAP improves from 55.54% to 67.38%.

We studied a novel re-ranking framework called RAE that exploits the information of keyword exemplars to complement the conventional ASR-based KWS systems.

Bibliographic details on Improved spoken term detection with graph-based re-ranking in feature space.

Graph-based re-ranking was proposed as a score calibration method in [4, 5] and has shown a considerable improvement in KWS retrieval performance on Babel ...

Lin-Shan Lee, “Improved spoken term detection with graph-based re- ranking in feature space,” in ICASSP, 2011. [10] Yun-Nung Chen, Yu Huang, Ching-Feng Yeh ...