Modelling F0 Dynamics in Unit Selection Based Speech Synthesis

Tihelka, Daniel; Matoušek, Jindřich; Hanzlíček, Zdeněk

doi:10.1007/978-3-319-10816-2_55

Cited by 5 publications

(1 citation statement)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to discriminate the segments coming from the corpus that fit the requirements expressed via the target sequence, the usual method [5] is to rank the units by evaluating the context matching degree (target cost) and the risk of creating an artefact if concatenating the unit (concatenation cost) via balanced cost functions. The concatenation cost typically relies mainly on acoustic features (MFCC, F0) [10,11] to evaluate the level of spectral resemblance between two voice stimuli on and around the concatenation point. As for now, concatenation costs are far from being perfect and audible artefacts appear both in commercial and research TTS systems, even after post-concatenation processing.…”

Section: Introductionmentioning

confidence: 99%

On the Suitability of Vocalic Sandwiches in a Corpus-Based TTS Engine

Guennec

Lolive

2016

Interspeech 2016

View full text Add to dashboard Cite

Unit selection speech synthesis systems generally rely on target and concatenation costs for selecting the best unit sequence. The role of the concatenation cost is to insure that joining two voice segments will not cause any acoustic artefact to appear. For this task, acoustic distances (MFCC, F0) are typically used but in many cases, this is not enough to prevent concatenation artefacts. Among other strategies, the improvement of corpus covering by favoring units that naturally support well the joining process (vocalic sandwiches) seems to be effective on TTS. In this paper, we investigate if vocalic sandwiches can be used directly in the unit selection engine when the corpus was not created using that principle. First, the sandwich approach is directly transposed in the unit selection engine with a penalty that greatly favors concatenation on sandwich boundaries. Second, a derived fuzzy version is proposed to relax the penalty based on the concatenation cost, with respect to the cost distribution. We show that the sandwich approach, very efficient at the corpus creation step, seems to be inefficient when directly transposed in the unit selection engine. However, we observe that the fuzzy approach enhances synthesis quality, especially on sentences with high concatenation costs.

show abstract

Section: Introductionmentioning

confidence: 99%