Cross-site combination and evaluation of subword spoken term detection systems

Abstract

The design and evaluation of subword-based spoken term detection (STD) systems depends on various factors, such as language, type of the speech to be searched and application scenario. The choice of the subword unit and search approach, however, is oftentimes made regardless of these factors. Therefore, we evaluate two subword STD systems across two data sets with varying properties to investigate the influence of different subword units on STD performance when working with different data types. Results show that on German broadcast news data, constrained search in syllable lattices is effective, whereas fuzzy phone lattice search is superior in more challenging English conversational telephone speech. By combining the key features of the two systems at an early stage, we achieve improvements in Figure of Merit of up to 13.4% absolute on the German data. We also show that the choice of the appropriate evaluation metric is crucial when comparing retrieval performances across systems

Subword-based automatic lexicon learning for Speech Recognition

“…This, in turn, consists of two parts: first, a pronunciation is hypothesized using phonemic subword units, and second, said pronunciation is converted to a spelling. Only generating a pronunciation for an unknown word is sufficient in applications such as Spoken Term Detection (STD) [6], where phonemic representations of speech are adequate for indexing and search. For transcription, however, an orthography needs to be estimated from a given phonemic subword sequence, as in [7], where phone transcriptions are converted to spellings using memory-based learning for Dutch OOV words.…”

Query-by-Example Spoken Term Detection ALBAYZIN 2012 evaluation: overview, systems, results, and discussion

“…Despite the bad performance exhibited by the configuration 4a corresponding to system 4, it must be noted that this was not optimized for the final metric (i.e., ATWV) but to get a predefined hit coverage, which greatly affects the final ATWV performance [62] and hence, a fair comparison with the rest of the systems cannot be made.…”