Topic
Naturalness
About: Naturalness is a research topic. Over the lifetime, 1305 publications have been published within this topic receiving 31737 citations.
Papers published on a yearly basis
Papers
More filters
01 Jan 2012
3 citations
TL;DR: In this article, the authors present an analogue space-time model for two-component BECs subject to laser-induced transitions between the components, and show that this model is an example for Lorentz invariance violation due to ultraviolet physics.
Abstract: So called "analogue models" use condensed matter systems (typically hydrodynamic) to set up an "effective metric" and to model curved-space quantum field theory in a physical system where all the microscopic degrees of freedom are well understood. Known analogue models typically lead to massless minimally coupled scalar fields. We present an extended "analogue space-time" programme by investigating a condensed-matter system - in and beyond the hydrodynamic limit - that is in principle capable of simulating the massive Klein-Gordon equation in curved spacetime. Since many elementary particles have mass, this is an essential step in building realistic analogue models, and an essential first step towards simulating quantum gravity phenomenology. Specifically, we consider the class of two-component BECs subject to laser-induced transitions between the components, and we show that this model is an example for Lorentz invariance violation due to ultraviolet physics. Furthermore our model suggests constraints on quantum gravity phenomenology in terms of the "naturalness problem" and "universality issue".
3 citations
23 May 2022
TL;DR: K-Converter is presented, a simple system to disentangle the timbre, pitch, and content information without any manual supervision or adversarial training and performs well in both speech naturalness and timbre similarity, with better robustness to comparisons.
Abstract: Singing voice conversion (SVC) converts a singer’s voice to another one’s voice while preserving the linguistic content. Recently, some SVC systems rely on supervised phonetic features extracted from pre-trained automatic speech recognition (ASR) models, increasing system complexity. Some end-toend SVC systems use adversarial training, which causes instability during optimization. To address these issues, we present K-Converter, a simple system to disentangle the timbre, pitch, and content information without any manual supervision or adversarial training. First, low quefrencies of mel-frequency cepstral coefficients (MFCC), which remove the glottal excitation mainly, are used as input representations. And the pitch-shift augmentation is used for further disentangling the pitch. Second, an encoder network is carefully designed to construct an information bottleneck, which learns to break up the pitch and timbre information of the source. Third, the content consistency loss is introduced to keep the content consistent between encoder outputs of source utterances and reconstructed ones. Experimental results show that our proposed system performs well in both speech naturalness and timbre similarity, with better robustness to comparisons.
3 citations
TL;DR: In this article , the authors used semi-structured interviews to study the effect of using physical food stimuli products on the perception of food products in the context of mothers and adolescents in terms of naturalness, sustainability and healthiness.
3 citations
TL;DR: This article used functional magnetic resonance imaging (fMRI) to test speech detection and language representation in the dog brain and found that bilateral auditory cortical regions represented natural speech and scrambled speech differently; with a better classifier performance in longer-headed dogs in a right auditory region.
3 citations