scispace - formally typeset
Search or ask a question
Topic

Naturalness

About: Naturalness is a research topic. Over the lifetime, 1305 publications have been published within this topic receiving 31737 citations.


Papers
More filters
Posted Content
TL;DR: This work proposes a novel approach that exploits concepts from non-standard analysis to introduce a notion of micro- and macro-steps in an extension of the TRIO metric temporal logic, called X-TRIO, to provide a formal semantics and an automated verification technique to Stateflow-like notations used in the design of flexible manufacturing systems.
Abstract: Many systems include components interacting with each other that evolve with possibly very different speeds. To deal with this situation many formal models adopt the abstraction of "zero-time transitions", which do not consume time. These however have several drawbacks in terms of naturalness and logic consistency, as a system is modeled to be in different states at the same time. We propose a novel approach that exploits concepts from non-standard analysis to introduce a notion of micro- and macro-steps in an extension of the TRIO metric temporal logic, called X-TRIO. We use X-TRIO to provide a formal semantics and an automated verification technique to Stateflow-like notations used in the design of flexible manufacturing systems.

3 citations

Posted Content
TL;DR: In this paper, a GMM-based mixture density network (MDN) was proposed to generate phone-level prosody patterns for multi-speaker TTS using speaker adaptation transforms of Gaussian means and variances.
Abstract: Generating natural speech with a diverse and smooth prosody pattern is a challenging task. Although random sampling with phone-level prosody distribution has been investigated to generate different prosody patterns, the diversity of the generated speech is still very limited and far from what can be achieved by humans. This is largely due to the use of uni-modal distribution, such as single Gaussian, in the prior works of phone-level prosody modelling. In this work, we propose a novel approach that models phone-level prosodies with a GMM-based mixture density network(MDN) and then extend it for multi-speaker TTS using speaker adaptation transforms of Gaussian means and variances. Furthermore, we show that we can clone the prosodies from a reference speech by sampling prosodies from the Gaussian components that produce the reference prosodies. Our experiments on LJSpeech and LibriTTS dataset show that the proposed method with GMM-based MDN not only achieves significantly better diversity than using a single Gaussian in both single-speaker and multi-speaker TTS, but also provides better naturalness. The prosody cloning experiments demonstrate that the prosody similarity of the proposed method with GMM-based MDN is comparable to recent proposed fine-grained VAE while the target speaker similarity is better.

3 citations

Journal ArticleDOI
18 May 2022-Forests
TL;DR: In this article , the authors developed a method to assess the quality of mountain forests using an index system composed of stand structure, site conditions, and landscape aesthetics at three criteria levels.
Abstract: Spurred by the degraded forest in the 2022 Chongli Winter Olympic Games area, the Chinese government initiated a national program for mountain forest rehabilitation. We developed a method to assess the quality of mountain forests using an index system composed of stand structure, site conditions, and landscape aesthetics at three criteria levels. The method involves index weights determined by the analytical hierarchy process (AHP) and entropy method. The results show that landscape aesthetics was the most important measure for the criterion layer. Slope aspect and naturalness were the most and second-most important indices, respectively, for the alternative layer. The quality of the mountain forest in the Chongli area was divided into four grades. The area had 7.8% with high quality, 46.7% with medium quality, 36.6% with low quality, and 8.9% with inferior quality. In total 76.6% of the damaged forest were distributed on sloping and steep sloping ground at 1700 to 2050 m altitude, and Betula platyphylla Sukaczev and Larix gmelinii var. principis-rupprechtii (Mayr) Pilg. were the predominating trees. The damaged forest was divided into over-dense, over-sparse, degraded, inappropriate tree species, and inferior landscape forest. For different types of damaged forest, corresponding modification measures were proposed. The methods developed in this study can be used for rehabilitation projects to improve the quality of degraded forests in mountainous temperate areas.

3 citations

Journal ArticleDOI
TL;DR: In this paper, audience ratings of the "naturalness" of spoken and written sentences were surveyed and found to be highly correlated with the naturalness of the sentences' utterances.
Abstract: (1966). Audience ratings of the “naturalness” of spoken and written sentences. Speech Monographs: Vol. 33, No. 2, pp. 156-159.

3 citations

01 Jun 2011
TL;DR: This paper argued that verbal innovation can be accounted for on the grounds of Natural Morphology, proposing that children opt for cognitively simple, i.e. unmarked derivational patterns, and come up with iconic coinages that are either metaphorical (in the sense of natural morphology), or diagrammatic and, at the same time, morphotactically and semantically transparent.
Abstract: This paper constitutes an attempt to shed some light on the formation of innovative verbs by Polish and English-speaking children. It argues that verbal innovation can be accounted for on the grounds of Natural Morphology, proposing that children opt for cognitively simple, i.e. unmarked derivational patterns. Consequently, they come up with iconic coinages that are either metaphorical (in the sense of Natural Morphology), or diagrammatic and, at the same time, morphotactically and semantically transparent. As might be easily predicted, iconicity manifests itself in the two languages in question through the use of different derivational mechanisms stemming from typological divergences between English and Polish.

3 citations


Network Information
Related Topics (5)
Statistical model
19.9K papers, 904.1K citations
69% related
Sentence
41.2K papers, 929.6K citations
69% related
Vocabulary
44.6K papers, 941.5K citations
67% related
Detector
146.5K papers, 1.3M citations
67% related
Cluster analysis
146.5K papers, 2.9M citations
66% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
2023282
2022610
202182
202063
201983
201852